RS 1803 HFS BP Enterprise Artificial Intelligence (AI) Services 2018

The Services Research Company
HfS Blueprint Report
Enterprise Artificial Intelligence (AI) Services 2018
March 2018
Tom Reuner
SVP, Intelligent Automation & IT Services
tom.reuner@hfsresearch.com
29j916d5
Table of Contents
Topic Page
Introduction and Key Definitions 2
Executive Summary 7
The State of the Enterprise Artificial Intelligence Market 10
Research Methodology 27
Service Provider Grid 32
Service Provider Profiles 38
Market Predictions and Recommendations 59
About the Author 66
© 2018 HfS Research Ltd. Proprietary │Page 1

29j916d5
Introduction and
Key Definitions
29j916d5
Introduction to and Scope of the HfS Blueprint Report:
Enterprise AI Services 2018 (1)
 The pace of change in the development and deployment of capabilities around the notion of Intelligent Automation (IA) and the
expansion toward artificial intelligence (AI) is nothing short of astounding. HfS remains at the forefront of engaging with
stakeholders around concepts, trends, market dynamics, and future directions. Many of those issues can be accessed through our
expansive research library on Intelligent Automation.
 AI is many things: It is hyped, it is undefined, it is becoming pervasive, and it is fostering emotional and at, times, heated discussions.
However, many of those discussions are more focused on consumer-facing issues such as self-driving cars, drones delivering Amazon
purchases, or robotic home helpers. The broader market is not yet recognizing the nearer-term impact of AI on B2B and Enterprise
operations. AI aims to automate intelligent activities that humans associate with other human minds through a combination of
reasoning, knowledge, planning, learning, natural language processing (communication), and perception (aka cognitive).
 While the broader market appears stuck in

discussions around RPA, we are keen to expand the
conversation toward notions of service orchestration
and integration around Intelligent Automation as
well as the impact of cognitive and AI solutions. A
useful context is the journey of organizations toward
the OneOffice. IA is critical to develop both the
Digital Underbelly as well as Intelligent Digital
Processes. Cognitive and AI in particular enable
those Intelligent Digital Processes to become
predictive as well as enabling. However, we
recognize that for some more IT-centric scenarios,
this model needs to adapted. Furthermore, while we
acknowledge the relevance of RPA, those
discussions are not in scope for this Blueprint
unless there are broad elements of AI capabilities
integrated.

29j916d5
Introduction to and Scope of the HfS Blueprint Report:
Enterprise AI Services 2018 (2)
 What we are trying to assess in this Blueprint is how service providers are orchestrating diverse sets of cognitive and AI solutions as
well as tools within the context of service delivery. How are they proactively transforming the processes for clients? The emphasis is
not on task automation or isolated point solutions, but on automation from a business function or process point of view. What are the
data management strategies underpinning the move to AI? Similarly, advisory is only relevant as part of implementations. These
capabilities might sit within traditional business units, but we are seeing the leading providers build out capabilities across those
business units.
The HfS Triple-A Trifecta
 As HfS has outlined in its Triple A Trifecta
framework, the Trifecta is where AI intersects with
Automation and Analytics. While each element of
the Trifecta has a distinct value proposition, there
is increasing convergence between the three
elements. For instance, smart analytics are
increasingly reliant on AI tools such as natural
language processing (NLP) to conduct search-
driven analytics, neural networks for data
exploration, and learning algorithms to build
predictive models. In fact, the Holy Grail of service
delivery transformation is at the intersection of
automation, analytics, and AI.
 The Trifecta is nonlinear without a definite
starting point. Transformation is not a linear
progression. Enterprises can start anywhere across The report is focused on
the Trifecta. It is not necessary to start with basic Artificial Intelligence
automation and then advance to AI-based (AI). Our definition of AI
automation. However, it is critical that you includes cognitive
understand the business problem that you are solutions.
trying to solve and then apply the relevant value
lever, or a combination of value levers.
29j916d5
AI Building Blocks (Illustrative, Not Exhaustive)
Technology: Nuance, Cortana, Alexa

NLP Use cases: voice recognition, conversational services
Machine and Technology: Google DeepMind, Tensorflow, Loop AI, Microsoft Cognitive Services
Deep Use cases: integration of data on an industrial scale, pattern recognition without ontology or knowledge-
Learning base, analysis of dark or IoT data
Neural Technology: HIRO, ignio, HOLMES

Networks Use cases: discovery in accounting, KYC, batch management
Technology: IBM Watson, Amelia, LivingActor, Accenture myWizard

Virtual
Use cases: substitute of IT and business agent, virtual data scientist, virtual scrum master, mortgage
Agents
broker advisor
Technology: IPcenter, HIRO, ignio, HOLMES

Autonomics
Use cases: password reset in IT help desk, self-healing IT environment
Computer Technology: AntWorks, AlchemyAPI, Clarifai

Vision Use cases: pattern recognition in images, integration of handwriting

29j916d5
HfS Definition of Intelligent Automation and AI Services
Intelligent Automation Value Chain
Plan Implement Manage Operate Optimize

• Advisory on • Program management • Governance • Infrastructure • New feature value
autonomics, cognitive for process automation management management identification and
computing, and AI • Process automation • Maintenance of • Application benefit analysis
• Workshops on IA and customization automated processes management • Ongoing adds and
vendor landscape and • Solution and technical • Optimization of BPO • IT help desk upgrades, migrations,
implications design contracts and SSC management and consolidation
• Automation • Process recording, delivery • BPO • Integration of big data
opportunity assessment mapping, and updating • Upgrade support analytics, and insights
• (Ro)Bot-as-a-Service
• Business case • Data extraction from • IA help desk • Best practice
• Real time analytics
development for heterogeneous systems understanding,
• Ongoing integration • Identify any required
automation documentation, and
• Leverage repository of • Support and changes in service
deployment end-user adoption,
pre-built components maintenance delivery or process to
• Operating model and utilities content creation, and
• Testing and QA account for changing curation
evaluation
• Predictive analytics business requirements
• Automation roadmap • New release and • User community
• Specialist development (e.g., M&A, divestment,
upgrade coordination participation.
• Compliance and risk modules new investments in IT)
assessment • Training and
• Enterprise systems • Mandatory regulatory
certification
• Security implications integration. adjustment ramification
• Acceptance testing management and
• HR and talent
management strategy • Change management. resolution.
• Governance policy
• Rollout strategy.

29j916d5
Executive Summary
29j916d5
Executive Summary (1)
 Enterprise AI services are one of the fastest growing market segments with annual growth in excess of 100%, albeit
this growth comes from a low base. As Enterprise AI is not one market, sizing the total market would be misleading.
Just for the RPA extension through AI, HfS estimates revenues of $1.6bn in 2018.
 AI is not one market, but AI should be seen as a set of technologies and building blocks. AI is not one market. AI not
only intersects with automation and analytics as the Triple A Trifecta framework suggests, but AI should be seen as a
set of technologies and building blocks that span a continuum and should be discussed within the context of business
operational impact, service delivery capability, and specific use cases. However, unlike Intelligent Automation, AI is
largely not plug-and-play.
 Enterprise AI is still at the periphery of the enterprise or applied as a bolt-on. Reflecting the early development
phase, organizations are pursuing AI solutions with a “bolt-on” approach, applying AI at the edge of the enterprise. As
the market matures, AI has the potential to disrupt and replace enterprise architectures and enterprise software.
 The Enterprise AI market has a duplexity of approaches: Industrialization and project-centric. We have to be
cognizant of the disparate starting points and context for projects. To get a better sense of the capabilities of service
providers, we need more differentiation in the discussions around AI. RPA and chatbots are low level; compare those
with the expansion of data science projects, autonomics, or even virtual agents, which have a high complexity and
require significant investments.
 The market is awash with PoCs predominantly around RPA extension, autonomics, and conversational services.
Being largely at the stage of strategy formulation, PoC or pilot stage means that organizations have yet to deal with
the commercial implications of deploying AI. Anecdotal evidence and perceptions of high levels of investment
necessary for AI projects remain critical inhibitors as organizations. This includes expectation that disruption caused
might offset cost savings.

29j916d5
Executive Summary (2)
 Hype around chatbots is distorting the marketing communications. For many buyers, all those chatbot PoCs are a
risk-free toe-dip into the sea of AI with the hope of learning some valuable lessons. However, many of those PoCs are
done without focusing on the customer by just offering narrowly defined conversations. The true value of
conversational services is being realized where a customer interaction is taken all the way to execution. The strategic
question: Who is providing those services? Will they be dedicated projects or will they be platforms such as Salesforce
or ServiceNow that are integrating lower level conversational services?
 The Holy Grail of AI is at the intersection of iterative data inputs and minimal training of algorithms. There is the
misguided expectation that one has only to throw machine learning at data and that would be sufficient to integrate
those data sets in production. Rather, organizations have to move to a data-centric mindset where data is the
centrepiece of digital strategies. This requires a new set of talent that blends the skills of data scientists and data
engineers. The focal point is reducing the time to train algorithms while being able to integrate iterative data inputs of
increasingly semi- and un-structured data. New technologies such as Gluon and AutoML will accelerate this process.
 Adoption of AI is across all verticals. While BFSI has the strongest investments, we are seeing deployments across all
verticals. Threshold setting for AML is a representative use case for complex data requirements. Another example of
the increasing complexity of projects is Hadoop cluster integration to feed every customer channel. Capabilities
include sentiment analysis and real-time analysis. Those use-cases are examples for scale and service orchestration
beyond the chatbot hype.
 The center of gravity for AI is around the mega-ISVs: Microsoft, Google, AWS, and IBM. Despite the plethora of AI
start-ups, the big bets and investments are being placed by the mega-ISVs. The best way to understand the evolving AI
supplier and capabilities landscape is to recognize the interplay of computing power optimized for AI, an expansive set
of algorithms, platforms enhanced for the particular use-case, and, lastly but most importantly, access to vast data
sets including unstructured data. Microsoft, Google, AWS, and IBM Watson are emerging as the central AI ecosystems
that will evolve into alliances such as Oracle and SAP in the ERP era.
 Service provider landscape. We have assessed 18 service providers across multiple dimensions on Enterprise AI
execution and innovation. The HfS Winner’s Circle for the 2018 Enterprise AI Services is comprised of IBM, Accenture,
TCS, Deloitte, Cognizant, and Genpact. Differentiating criteria are the variety, depth, and scale of use-cases as well as
the robustness of delivery.
29j916d5
The State of the Enterprise
Artificial Intelligence (AI)
Market
29j916d5
AI Is Progressing at an Astounding Speed But Success Will
Ultimately Depend on the Business Impact It Creates
 The pace of change is nothing short of astounding: Providers are calling out an unprecedented speed of change with
engagements moving from prototype to project in two to three months. At the same time, stakeholders are seeing an
explosion of the automation eco-system. While the majority of stakeholders doesn’t expect disruption from the start-
ups because R&D costs too high in order to disrupt entire ecosystems, some indicated that this explosion of
companies and capabilities is making them rethink the cloud stack and other broader assets.
 Enterprise AI is still nascent, yet projects already dwarf RPA: Despite all the marketing noise, there is little guidance
and education on Enterprise AI. Much of the noise is around consumer facing applications such as Siri and Alexa as
well as around the socio-economic impact such as job losses, the need for a basic income, or even fear that AI could
lead to the destruction of society. While the market is still nascent and broadly speaking under the radar, the scale,
depth, and value of the AI deployments dwarf anything that we are seeing in RPA. In many respects, HfS feels
reminded of the early RPA days as we have to explain the basics to buyers and guide them with frameworks as well as
lessons learned from the early deployments.
 The market is characterized by a duplexity of AI: As slide 13 highlights, the Enterprise AI market can be seen as a
duplexity of approaches: both industrialization and project-centric approaches. While the boundaries of these
approaches are fluid, industrialization is all about finding as many commonalities across delivery backbones as
possible in order to scale and save costs at the same time. The strategic logic is 1-to-many. Examples would be
monitoring of infrastructure or self-remediation as technologies including IPsoft and Arago. Conversely, project-centric
approaches are highly domain-specific and the strategic logic is 1-to-1. Examples include the automation of a medical
coding in a hospital to support diagnoses and improved patient records.
 Clients don’t buy technology; they buy outcomes. Boards are paranoid about automation and AI, but they struggle to
turn it into an actionable mandate. As a result there is a lot of experimentation by process owners and program
managers. Yet, fundamentally they don’t want to buy technology change agents such as RPA, AI, or Blockchain off the
shelf, rather, they want to procure an outcome. It is here where we see a disconnect with the supply side, which for
organizational reasons (including sales incentives) continues to evangelize on technology issues.

29j916d5
AI Is Not One Market, But AI Should Be Seen As a Set of
Technologies and Building Blocks
 Enterprise AI is a segment with many Just like Intelligent Automation, AI should be seen as a continuum
markets: AI is not one market. AI not only
intersects with Automation and Analytics as
the Triple A Trifecta framework suggests, but
AI should also be seen as a set of
technologies and building blocks that span a
continuum and should be discussed within
the context of business operational impact,
service delivery capability, and specific use
cases. As such, AI discussions should follow
the framework that HfS
has developed across the broader
notion of Intelligent Automation.
While nothing in the context of automation

is commonly defined, the HfS Intelligent
Automation Continuum is aiming to provide
stakeholders with a point of reference to
discuss how innovation around automation
is impacting enterprise operations.

29j916d5
The Enterprise AI Market Has a Duplexity of Approaches:
Industrialization and Project-Centric
Industrialization Project-Centric
Service orchestration Specific requirements
Horizontal Sub-sector lens
Out-of-the-box Design Thinking

Alignment with Intelligent Expansions analytics
Automation
Data Science
Service delivery
Proprietary IP and
Mega ISVs open source
Data lake Data silos
RPA, autonomics, Proprietary algorithms,
chatbots Deep Learning
Narrow AI Strong AI
Industrialization is all about finding as many Project-centric approaches are highly domain-specific and
commonalities across delivery backbones as possible the strategic logic is 1-to-1. Examples include the
in order to scale and save costs at the same time. automation of a medical coding in a hospital to support
The strategic logic is 1-to-many. Examples would be diagnosis and better patient records.
monitoring of infrastructure or self-remediation
technologies including IPsoft and Arago.
29j916d5
Context, Context, Context: We Have to Be Cognizant of the
Disparate Starting Points for AI Projects
The AI journey has disparate starting points
 Suggestions of explaining AI better as
“augmented” intelligence are just
semantics. The fundamental point is about
what business problems we can solve, not
about getting confused by movies like The
Matrix and discussion about the Singularity.
The key question is: What can’t scientists
do today with traditional approaches?
 Put in other words, we have to be
cognizant of the disparate starting points
and context for projects. Not least in order
to get a better sense of the capabilities of
service providers, we need more
differentiation in the discussions around AI.
 RPA and chatbots are low level; compare those with the expansion of data science projects, autonomics, or even virtual agents,
which have a high complexity and require significant investments.
 For many buyers, all those chatbot PoCs are a risk-free toe-dip into the sea of AI with the hope to learn some valuable lessons.
However, how do you compare this to the Watson projects costing millions trying to cure world hunger? In summary, AI is not a
linear progression, neither from RPA nor chatbots nor any other starting points. This points to both opportunities but also
disruptive threats.
 There can be many other variations of those starting point, not least with a highly specific vertical context, but
this approach of clustering might help to convey the fundamental differences between those five clusters.
29j916d5
The Explosion of AI Start-ups Is Resulting in Partnership
Ecosystems for Most Serious Enterprise AI Service Providers
Virtual
Agents Neural
Networks
NLP
AI Building Machine/
Blocks Deep
Learning
Computer
Vision Autonomics

29j916d5
The Center of Gravity for AI Is Around the Mega-ISVs:
Microsoft, Google, AWS, and IBM
 Despite the plethora of AI start-ups, the
big bets and investments are being
placed by the mega-ISVs.
 The best way to understand the evolving
AI supplier and capabilities landscape is
to recognize the interplay of computing
power optimized for AI, an expansive set
of algorithms, platforms enhanced for
particular use-cases, and lastly but most
importantly, access to vast data sets
including unstructured data.
 While the direction of travel is toward
autonomous processes, we need people
developing, improving, and refining the
AI, its environment, and necessary
algorithms.
 The use-case will determine the best
technical approach, but critically will
have to leverage commonality across
their delivery backbones and develop
the capabilities to route customer data
through that backbone from interaction
all the way to execution or conclusion.
 Thus, AI is reinforcing the necessity for
organizations to drive service
orchestration strategies.

29j916d5
Despite the Hype, the Current Enterprise AI Market Is in a
State of Experimentation with Low-Level PoCs Around
Chatbots and RPA Extensions
 Low-level PoCs around chatbots and RPA For which processes have you implemented AI Automation?
extension projects that are the majority of
the AI deployments reference the nascent Customer Service 46%
state of the market.
 This is reinforced by our recent study titled Finance & Accounting 41%
State of Automation 2017 that customer
IT including service desk 39%
service and F&A are the main processes for
which AI has been deployed. Those Procurement 37%
projects typically focus on FAQs and next-
best action for customer service as well as Master Data Management 36%
the integration of Google and Microsoft
cognitive libraries into RPA deployments. HR 35%
 While boards are paranoid about the Supply Chain 34%
impact of AI, they struggle to turn it into
actionable agendas. Thus, few clients go for Industry specific processes 33%
transformational projects. Rather, we are
seeing narrow AI approaches around lower Marketing 27%
level chatbots and RPA extension that
dominate the market.
Source: HfS Research “State of Automation 2017”
Sample: n=181 Buyers of AI

29j916d5
Innovative AI Use-Cases Are Starting to Emerge,
But They Are Rare
Virtual Assistant integrated with Hadoop cluster: Big 4 for global bank, integration of Hadoop
cluster to feed into every customer channel; included sentiment analysis and real-time
analysis. Example for scale and service orchestration; goes beyond chatbot hype.
Exploring quantum computing techniques on AI and Machine Learning

algorithms in Blockchain applications: Global SI for leading bank in Australia.
Example for integrated approach of next-gen technologies and approaches.
Setup of joint AI CoE with Intel and support of Intel Nervana AI Academy. TCS provides a
platform to connect researchers, developers, data scientists, and startups. These solutions will
be optimized for the Intel portfolio of technologies Example for ecosystem enablement.
Leverage of ML for medical coding: Global SI (not Watson) helped European

hospital to apply medical coding at scale to allow for digital patient record and
diagnosis. Example for complex Data Science approach at scale for critical processes.
Integration of disparate sources for General Ledger: Global SI. Clients can drop
disparate information for General Ledger requirements; Machine Learning and other
technology building blocks allow for seamless processing. Example for reimagining processes.
Cross-fertilization from other sectors: Global SI has helped Australian company to

automatically identify telephone posts using Google TensorFlow and Street View
replacing manual inspection; broad replicability for other sectors, think insurance scenarios.

29j916d5
The Nascent Nature of Enterprise AI Is Characterized by
Demand-Side and Supply-Side Challenges
Supply-Side Challenges (1)
 Lack of organizational maturity around AI: The nascent state of the market is also referenced by the fact that only few providers
are moving toward a formalized AI practice or building practices around AI-specific partners. Thus far, only Accenture and KPMG
have moved to a formalized AI practice. For most providers, AI remains a critical building block in their Intelligent Automation
strategies. While we are seeing a slew of IBM Watson practices springing up, capabilities around Google, Microsoft, and AWS are
part of broader alliance programs rather than dedicated AI programs. At the same time, providers like Cognizant work with an
incubator approach, in this case around “conversational services”. Only IBM, Accenture, and KPMG have formalized AI practices,
while most providers drive AI from their analytics business units. Cognizant is using an incubator model, however, with a strong
emphasis on conversational services. Broadly speaking, the success of AI projects currently is often left up to individuals and
teams rather than organizational structures and strategic initiatives.
 No mainstream organizational model around data yet: It is unclear as yet what will become of the mainstream organizational
model around data issues. Is it an emphasis on the quantity of data such as linking data lakes with AI capabilities? An emphasis on
quality such as sub-sector focus? Or rather a federated model blending different approaches? Depending on the size of a provider,
a federated model appears most likely. Both the lack of organizational maturity and, more importantly, the not yet aligned data
management strategies are curtailing the scale of AI projects. Regarding the alignment of data management strategies we are
seeing different schools of thought in the market. Providers with a strong emphasis on industrialization are aiming to integrate
expansive sets of data lakes with their AI capabilities. A good example for this is IBM’s Services Platform with Watson. Conversely,
EY, with a much more domain-specific approach, suggests “most of the data needs ought to be applied with a sub-sector lens”. Or
put in other words, how much data do you need?
 Scarcity of talent: Despite all the marketing noise, the talent for AI capabilities is scarce. For instance, data scientists are a rare
and expensive breed. It will take time for universities to adapt to the fast changing market. Crucially, it is not just about technical
capabilities, as one senior executive quipped colourfully, "Data scientists can do the technology, but they don't have practical
experience […] IBM, SAP all those guys, no one has enough resources in this space. Hence they roll out all those professors
without any delivery experience.“ But it also flips in the other direction as another executive pointed out, "You easily drown in
your own success, hence you have to plan accordingly”.

29j916d5
 Platform approach as response to scarcity of talent: Platforms like Salesforce Einstein, SAP Leonardo, or innovative technology
providers like CognitiveScale, which offers pre-trained machine learning capabilities and curated data sets, are addressing the
skills shortage. However, the key strategic question is how far such a “standardized” approach can deal with the complexity of
organizations processes. The strategic intent is to make AI accessible. The focal point is the democratization, commoditization,
and templating of the work of data scientists as well as training algorithms.
 AI is rarely plug-and-play: Despite the noise around platforms, the biggest value of AI is where it is leveraged to re-imagine
processes rather than where it is used for industrialization. As one executive put it “the industry has to move an AI-patch
model to and AI-led model”. Another way of looking at it is that, for instance, managed services are AI supported whereas
projects tend to be AI led. As such, AI is neither the linear progression on the Intelligent Automation Continuum nor does it
lend itself to expansive plug and play. It is rather about moving to a data-centric mindset.
 Immature commercial models: As the market is awash with PoCs, it is not surprising that we lack robust insights into the
commercial implications of AI. On the transactional side, organizations still struggle to move away from FTE and AHT-centric
models. Yet, a key issue is clarity on ownership and accountability for data. Organizations should look for opportunities to use
machine learning techniques to tap into “dark data”—information that organizations collect during their regular business
activities but do not currently use. At the same time, buyers are wary of vendor lock-in, therefore data portability will emerge
as a critical differentiator.
 Lack of risk considerations: Aligned with the immature commercial models, the lack of risk consideration is striking. Not least
as AI is meant to be a conduit to achieve end-to-end processes. As one sell-side executive put it: “The question how do we
manage risk never comes up; yet elevated risk levels are one of the most crucial issues on the journey toward AI”.

29j916d5
 Focal point knowledge: Knowledge is critical for the institutionalization of AI. However, the investments are around innovative
approaches such as knowledge graphs that allow the leverage of data assets across an organizations similar to the way
Facebook is dealing with information and data. Yet, as executives from Capgemini pointedly put it, knowledge management
systems have become a corporate landfill across the industry. Put in other words—this is not old wine in new bottles.
 Automation and AI won’t lead to a reduction of the strategic importance of offshoring: Against some bold statements that
automation will lead to the death of offshoring, we are expecting a shift in capabilities rather than a diminished strategic
importance. AI is no different from the broader notion of automation. Take the example of Accenture: 70% of its AI talent sits
in India and a further 20% in APAC.

29j916d5
Demand-Side Challenges (1)
 Buyers are confused about AI: As with most things around Intelligent Automation, AI is neither commonly defined nor are
providers clear about the business challenges addressed or able to offer a clear depiction of cases. As we have suggested,
boards are paranoid about the impact of AI; they struggle to turn it into actionable agendas. Thus, middle management is
feeling the pressure to drive innovative approaches as they have precious little guidance from industry stakeholders about
what their mandate should be and what lessons were learned from early deployments. In many respects, it feels like
Groundhog Day—a reminder of the early days of RPA where basics and the framework need to be explained to industry
stakeholders. As a direct consequence, for instance, TCS executives suggested that they don’t talk about AI in client
discussions, they instead tend to focus on the business problems.
 Focus on technology versus “outcomes”: Buyers are clear in that they don’t buy RPA, AI, or Blockchain off the shelf; they
instead aim to procure an outcome. However, executives at EY rightly pointed out, “It is way too early to sell outcomes
around AI”. Therefore, we rather should understand those outcomes as business problems and use cases.
 Scarcity of talent: The talent that understands all the innovations subsumed by the AI moniker and, most importantly, their
impact on process chains and workflows is scarce. The progression toward a data-centric model requires a blend of data
engineers and data scientists to harness the value of data. And, more broadly, data has to become the corner stone of an
organization’s digital strategy, not just a by-product. Platforms such as Salesforce Einstein and SAP Leonardo are trying to
mitigate the dependency on such high level talent. Organizations have to focus on evaluating processes where a platform
approach can address their requirements. As such, it is similar to RPA. However, it will take significant time for the platform
to mature. As with all things around AI, the approaches will become more robust the more algorithms are being trained.
 Lack of alignment between data and delivery strategies: As the supply side is still unsure as to what the mainstream model
of aligning data with AI technology building blocks might become, buy side organizations should seek advise from
consultancies like the Big Four. However, the more fundamental challenge is to translate such top level guidance into
managing day-to-day operations. There will be a new set of job functions emerging that organizations have to focus on.

29j916d5
Demand-Side Challenges (2)
 Governance, testing, and security hardly addressed: Which of the following would AI Automation have to be able to
Given the lack of education and communication, it demonstrate before your organization considers using it?
might be not surprising that as an industry we have
hardly touched discussions on governance, testing,
and security. However, the more we progress toward
notions of self-learning and self-remediation, the
more urgently we need discussions how the
operating model for digital is being transformed and
challenged by the rise of AI. Yet, as a buy-side
executive reflected self critically, “We didn't know
our processes well enough and they weren't properly
documented”. Thus, both supply and buy side have
to engage in new ways of collaboration. Another
executive added that such a collaboration has to look
for new model, “Our governance is too heavy for
Intelligent Automation; we need more agility“.

Sample: n=219 Non-Buyers of RPA

29j916d5
Moving Toward a Data-Centric Mindset Necessitates
New Requirements for Talent
 The Holy Grail of AI is at the

intersection of iterative data inputs
Data Scientist: Data Engineer: and minimal training of algorithms.
• Cleaning and organizing • Designing, testing,
data maintaining scalable  To drive scale and industrialization,
• Custom algorithms data architectures technology providers like
• Statistical modelling • Evaluation, CognitiveScale offer a blend of machine
• Feature engineering integration tools learning training, pre-built agents, and
• Exploratory • Data ingestion curated data sets to operationalize AI.
analysis • Deployment
• Solution architecture
Data  Those approaches are managing
handoffs between data scientists,
analysts, application developers, and
process owners. Similarly, service
providers are aiming at the
industrialization of use-cases.
Artificial Intelligence
Technologies:  Examples are Accenture’s myWizard,
• Ingestion of data Infosys’ Nia, and Atos’ Ystia platforms.
• Pattern analysis The intent is to offer a common
• Knowledge representation workplace to store, share, retrieve,
• Integration of disparate and update assets, models,
approaches and frameworks.

29j916d5
The Holy Grail of AI Is at the Intersection of Iterative Data Inputs and
Minimal Training of Algorithms
 Without access to vast data sets, AI will remain
narrow, thus carrying out only specific tasks.
Algorithms need more data than currently
available to progress toward general AI with the
goal to handle tasks from different areas and
origins, such as any intellectual task that a
human being can handle.
 The difference between Big Data projects and

the need for integrating data on an industrial
scale to progress toward the OneOffice is
perspective is moving the perspective from the
rear mirror view to the windshield. However,
the data traditionally held in data warehouses
cannot easily be leveraged for prescriptive
analytics utilizing machine learning.
 Machine learning requires “learnable” data.

That is, data that is known to contribute to
inference. Crucially, those algorithms providing
it need to be trained in order to leverage those
data sets in production. To progress toward
autonomous processes that are at the heart of
the OneOffice requires the access to
increasingly unstructured data. Only when
algorithms can learn from those vast data sets
will we inch further to the notion of General AI.
In simple terms, less human interaction is
required with more data that can be integrated
and analyzed.
29j916d5
Enterprise AI Will Have a Profound Impact on the Ideals of
the As-a-Service Economy
Strong Modest
29j916d5
Research Methodology
29j916d5
Required Information and Model Weighting
Execution
How the service provider works with clients to integrate automation into their delivery capabilities 25%
How the service provider addresses training and supporting clients 5%
Actual delivery of services 30%
Scale and repeatability of deployments 20%
How the service provider is leveraging service orchestration to deliver end-to-end and point solutions 10%
How the service provider supports clients with use case identification and match those with most relevant
tools and approaches 10%
Innovation
Vision for and investments in the evolution of intelligent automation 25%
Expanding service delivery toward AI 15%
Tool and platform strategy for intelligent automation delivery 20%
Data management strategies supporting the integration of semi and unstructured data 10%
Governance and testing services (e.g., services that offer innovative approaches for cognitive, AI, and self-
remediating engines) 10%
Leverage of Design Thinking, co-ideation for reimagining and transforming processes 10%
Integration of process and organizational consulting 10%
Note: These scores will be based on the RFI response and the feedback from the client references.

29j916d5
Execution Definitions
How well does the service provider execute on its contractual agreement, and how
Execution
well does the provider manage the client/provider relationship?
How the service provider works with clients
How engaged is the executive and management team in defining and managing the
to integrate automation into their delivery
delivery of IA? Do providers help clients to understand the end-goal of automation?
capabilities
Is the service provider addressing the impact on talent? How is the provider advising
How the service provider addresses training
on and supporting the transformation of knowledge work? Is there a structured
and supporting clients
approach to training?
What are the client’s and market’s overall impression of the quality of service across
Delivering actual services the value chain from this service provider? How is IA helping clients to transform
processes? What is the overall impact on client processes?
How is the service provider addressing scaling deployments across clients’ processes?
Scale and repeatability of deployments
Is the service provider seeking to move to repeatability of IA solutions and projects?
When looking at a client’s IA issues, can the service provider offer various solutions
How the service provider is leveraging
(point and end to end) to create a flexible and configurable (or customized) response?
service orchestration to deliver end-to-end
How is the service provider leveraging approaches for service orchestration and
and point solutions
integration to deliver IA?
How the service provider supports clients
with use case identification and matches How is the service provider engaging with clients to transform their processes? What
those with most relevant tools and is the starting point for those discussions?
approaches

29j916d5
Innovation Definitions
How well does the service provider innovate its offering(s) in response to market demand,
Innovation
client requirements and its own vision for how the IA market will evolve?
What is the service provider’s vision for the evolution of IA? Is there a clear strategy for
Vision and committed investments for
delivering broader capabilities As-a-Service and are there identifiable investments in place to
the evolution of intelligent automation
realize this strategy today?
How does the service provider expand its delivery capabilities toward AI? What are examples
Expanding service delivery toward AI for both partnerships as well as successful deployments? How do these deployments differ
from RPA projects?
What is the role of tools and platforms in the service provider’s offering strategy? Are the
Tool and platform strategy for
selected platforms developed in-house or are they provided by third parties? Is there a
intelligent automation delivery
demonstrable intent to maintain and enhance the in-house platforms?
How does the service provider work with clients to develop a comprehensive set of data and
Data management strategies then in turn make the analysis of that data and the development of insights possible? Is the
supporting the integration of semi and service provider allowed by clients to act upon the developed data to improve the
unstructured data effectiveness of overall service delivery? Does the service provider have a vision to integrate
IA with actionable data?
Governance and testing services (e.g.,
services that offer innovative How are these new environments to be governed? How is the service provider testing these
approaches for cognitive, AI, and self- new environments? How does this approach differ from traditional testing approaches?
remediating engines)
Integration of process and How does the service provider combine capabilities in process and organizational consulting
organizational consulting and supporting technology in the form to create innovative IA offerings?
Leverage of design thinking, co- How effectively does the service provider embed Design Thinking methodologies and notions
ideation for reimagining and of co-ideation in its IA engagements to define, prioritize, and execute against business
transforming processes outcomes of IA initiatives?

29j916d5
Research Methodology
Data Summary Participating Service Providers
 The data for this Blueprint was collected in Q3/4 2017,
and covers services buyers, service providers, and
advisors and influencers of Intelligent Automation.
 Invitations were sent to the top 20 service providers
and consultancies by revenue.
 This report builds on the research for the HfS Intelligent
Automation Blueprint 2017 and the AI-enabled
OneOffice Premier League 2017. Furthermore, it is
leveraging the ongoing discussion with stakeholders as
part of HfS Intelligent Automation practice.
This Report Is Based On:
 Tales from the Trenches: Interviews were conducted
with buyers who have evaluated service providers and
experienced their services. Some were supplied by
service providers, but many interviews were conducted
by HfS Executive Council members and participants in
our extensive market research.
 Sell-Side Executive Briefings: Structured discussions
with service providers were intended to collect data
necessary to evaluate their innovation, execution,
market share, and deal counts.
 Publicly Available Information: Financial data, website
information, presentations given by senior executives,
and other marketing collateral were evaluated.

29j916d5
Service Provider Grid
29j916d5
Guide to the Blueprint Grid
To distinguish service providers that show competitive differentiation in a particular line of delivery with
progress in realizing the As-a-Service Economy of business-outcome-oriented, on-demand talent and
technology services, HfS awards these providers the As-a-Service Winner’s Circle designation.
Execution Innovation
Winner’s Circle Collaborative relationships with clients, Articulates vision and a “new way of
shows excellence recognized by clients services executed with a combination of thinking,” has recognizable investments in
in the Eight Ideals in execution and talent and technology as appropriate, and future capabilities and strong client
innovation flexible arrangements feedback, and is driving new insights and
models
High Performers Execute some of the following areas with Typically describe a vision and plans to
demonstrate strong capabilities but excellence: worthwhile relationships with invest in future capabilities and
lack an innovative vision or clients, services executed with “green partnerships for As-a-Service and illustrate
momentum in execution against the lights,” and flexibility when meeting clients’ an ability to leverage digital technologies
vision needs or develop new insights with clients
High Potentials Early results and proof points from Well-plotted strategy and thought
demonstrate vision and strategy but examples in new service areas or leadership, showcased use of newer
have yet to gain momentum in innovative service models but lack scale, technologies or roadmap, and talent
executing the strategy and vision broad impact, and momentum in the development plans
capability under review
Execution Powerhouses Evidence of operational excellence; Lack of evident vision and investment in
demonstrate solid, reliable execution, however, still more of a directive future-oriented capability, such as skills
but have yet to show significant engagement between a service provider development, “intelligent operations,” or
innovation or vision and its clients digital technologies

29j916d5
HfS Blueprint Grid: Enterprise AI Services

29j916d5
Assessing AI Capabilities and Maturity
 AI, just like automation, is in the eye of the beholder. Therefore, we have to be cognizant of both
the different starting points and context of projects. The scale, complexity, and necessary
investments are vastly different for projects like (low level) chatbots and (high level) data science.
 Similarly, providers have a different strategic focus and sweet spots. System integrators and BPOs
are predominantly about industrialization services and offerings, while for instance the Big Four
consultancies have a strong domain-specific focus such as vertically aligned data science projects.
 Thus, the mapping on slides 36 and 37 is an additional way to provide more differentiation along
those focal areas and starting points for AI projects
 The criteria used in the mapping includes progress on the formalization of an AI practice as a
measure of organizational maturity and data-centric alignment as to how data and AI strategies are
being aligned and integrated. Furthermore, we try to offer more differentiation between low-level
projects like conversational services and RPA extension and more complex projects like data
science expansion and autonomics.
 This guidance is meant for organizations looking for insights who to include in their RFIs and
strategic initiatives.

29j916d5
AI Capability and Maturity Mapping
DXC.
Accenture Atos Capgemini Cognizant Deloitte technology EY Genpact HCL
AI practice (formalized)
Industrialization
Domain-specific offerings
Data-centric alignment
Extension RPA
Conversational services
Autonomics
Expansion data science
Machine learning
Strong = Most Maturity
Modest = Least Maturity

29j916d5
AI Capability and Maturity Mapping
Tech
IBM Infosys KPMG LTI PWC Syntel TCS Mahindra Wipro
AI practice (formalized)
Industrialization
Domain-specific offerings
Data-centric alignment
Extension RPA
Conversational services
Autonomics
Expansion data science
Machine learning
Strong = Most Maturity
Modest = Least Maturity

29j916d5
Service Provider Profiles
29j916d5
HfS Intelligent Automation Value Chain: Key to Profiles
 Value Chain coverage is indicated by blue shading.
 Grey shading indicates that the service provider does not offer these services.
For example:
Full Value Chain offered “Build” not offered
Design Design
Coverage Coverage
Methodologies Methodologies
Build Build
Run Run
 Each profile also includes an “Operations” box where we list service provider statistics.
 “N/A” indicates that the service provider does not have this service today. In many cases, the service provider
may have these certification applications in progress.

29j916d5
Major Service Provider Dynamics: Highlights
Execution Innovation
IBM is the leading provider in execution as clients cite best-in- Accenture is the leading provider in innovation. Its
class transformation capabilities that are underpinned by myWizard platform is a lighthouse investment for
expansive IP including Watson, Weather Channel, and innovation. Clients reference that Accenture doesn’t
AlchemyAPI. Clients value IBM as a strategic partner that has come across transactional; rather it comes across as a
demonstrated the robustness of automation without a single (or strong thought-leader with a more consultative
at least very limited) outage. approach: “They are making AI easy for us“.
▪ Actual delivery of services: The key reference points are the ▪ Vision for and investments in the evolution of
depth and scale of use cases along the duplexity of AI. Intelligent Automation: Given the duplexity of AI,
Furthermore, the focus is on solving business problems rather service providers are at the forefront of integrating AI
than engaging around technology capabilities. IBM, Accenture, into their delivery backbones. Accenture, IBM, TCS,
Deloitte, and TCS are standing out in this regard. and HCL stand out for their vision for the evolution of
Intelligent Automation and AI.
▪ Scale and repeatability: Invariably stronger developed on the
service provider side as they aim to industrialize service ▪ Expanding service delivery toward AI: Broadly
delivery. IBM, Accenture, and TCS are being called out by speaking, the market is early in expanding toward AI.
clients for blending AI into broader service provider But both the investments and pace of change are
capabilities. Deloitte and EY are referenced for robust RPA astounding. IBM, Accenture, Infosys, Atos, and TCS
extension, driving deployments as part of broader digital impressed both in terms of vision as well as
initiatives. capabilities.
▪ Supports clients with use case identification and matches ▪ Governance and testing services: Only few providers
those with most relevant tools and approaches: As was with have demonstrated thought leadership and best
the early RPA market, the identification of use cases and practices around AI. Accenture’s Teach and Test
matching those with tools and technologies is critical for solution is leading the market with addressing issues
achieving robust delivery. Accenture, IBM, and Deloitte are such as bias as well as taking the end-to-end processes
leveraging their consulting expertise to achieve tangible in mind. Similarly, IBM and TCS have demonstrated
results. both though-leadership as well as robust capabilities.

29j916d5
Accenture Global service provider focusing on innovation and transformation
Winner’s Circle
Blueprint Leading Highlights Strengths Challenges
• Vision for and investments • Focus on strategic and transformational projects with a view to create self-adapting, self-optimizing, “living • Balancing innovation with consistent levels of
in the evolution of IA processes” that use machine learning algorithms and real-time data to continuously improve. Clients reference delivery. Accenture’s focus on transformation and
• Expanding service delivery that Accenture doesn’t come across as transactional; rather it comes across as a strong thought leader with a innovation invariably raises the issue of talent for
toward AI more consultative approach, ”They are making AI easy for us“. those innovative services. As some staff is being
• Data management strategies • Organizational maturity tangible in formalized AI practice: Accenture has formed an Applied Intelligence group trained on projects, consistency in terms of
supporting integration of within Accenture Digital, bringing together the capabilities developed under the AI Strategic Growth Initiative capabilities and availability was raised as a
semi and unstructured data challenge by some clients.
with Accenture Analytics. The group will direct strategy, innovation, and market presence while embedding
• Tool and platform strategy
Applied Intelligence into every part of their business. This is being reinforced by Google, Watson, and Amelia • Change and stakeholder management: Some
for Intelligent Automation
delivery practices. business process-centric clients would like to see
• Strong traction for myWizard lighthouse offering: myWizard makes Accenture’s automation investments more proactive support for engaging with
Plan available in one place and serves as a platform for open innovation. With more than 2,300 client engagements stakeholders on change management, especially in
now on the platform, it has become an immense data pool that helps to train virtual agents, make the context of managing IT implications. The project
Implement recommendations, and act. As the data pool deepens, automation becomes more advanced and more reliable, teams did not cover those implications or have the
and software engineers can apply it more extensively for clients. appropriate skill sets to do so.
Manage • Breadth of vertical and domain-specific offerings: Leveraging the AI Navigator, a “menu” of AI solutions that • Mitigating perceptions of high price amidst hype of
are readily available to clients and accessed through the Accenture Intelligent Automation Platform. Accenture low-level chatbots and ubiquitous proof of concept,
Operate delivers process transformation at scale spanning multiple industries and functional areas. Examples include further exacerbated by formalized Watson and
Intelligent Email Advisor, Intelligent Invoice Advisor, Pharmacovigilance Advisor, and Claims Adjudication Amelia practice, which carry high price
Optimize Advisor. connotations.
Relevant Acquisitions/
Key Clients Operations Technology Tools and Platforms
Partnerships
Acquisitions include: Accenture works with clients Geographic footprint and scale of the AI practice: • Accenture myWizard: An enterprise automation platform that provides a
• Search Technologies across industry sectors: spectrum of automation services, including bots, integrators, analytics,
(2017) More than 6,000 AI professionals across a spectrum machine learning, deep learning applications, virtual agents, technology and
• Genfour (2017) • APAC utilities company of AI technologies and platforms, including natural industry assets and cognitive tools. It automates a range of activities including
• OPS Rules (2016) • Land Registry Department of a language processing, machine and deep learning, planning, ticket resolution, monitoring, issue prediction, agile delivery
• Furthermore, major European country neural networks, virtual agents, autonomics, and execution, requirements and user story analysis. It enables a common data
investments in MightyAI, • Major Italian government computer vision. fabric and integration of multiple proprietary and third-party AI solutions and
Bitvore, 1Qbit and public agency • About 3,000 are considered “deeply skilled and tools.
Paxata • US-based telecom company experienced” in bringing AI solutions to clients. • Accenture Intelligent Automation Platform: Platform integrates four
• Global oilfield services • About 1,000 use the myWizard platform to essential parts—Business Workflow Management, Delivery Management,
Partnerships include: company deploy AI and automation solutions for Intelligent Automation, and Analytics and Insights, with a neutral ERP
• AWS • Global insurer based in Europe application services clients. interface at the core, it drives seamless integration of external data sources
• Microsoft • US-based e-commerce • About 400 are dedicated to Accenture and client systems to accelerate decision making and business outcomes.
• Google company Operations where they develop AI use cases, • Accenture Insights Platform: A flexible analytics-as-a-service platform, built
• IBM • US fire and safety equipment product designs, and data science from an integrated suite of leading technologies that ingest, store, and apply
• Splunk major requirements. data to generate real-time, actionable intelligence. Leveraging AI and pre-
• Tricentis • Woodside • Approximately 70% of AI professionals are developed solutions, it speeds clients’ time to value.
• IPsoft • Avianca located in India and 20% through the rest of • Video Analytics Service Platform: Automatically identifies events related to
APAC, 5% in the Americas, and 5% in Europe. security and safety, operational efficiency, and customer insights.

29j916d5
Cognizant Global service provider with strong exposure in health sector,
Winner’s Circle blending data with human science
• Actual delivery of services • Eco-system mindset: Cognizant pursues a blend of proprietary and third-party tools. It has a structured • Progress toward holistic automation platform and
• Data management strategies approach to screen the innovation in Intelligent Automation and AI. Cognizant’s vendor-agnostic Smart foundational AI layer: Although Cognizant has broad and
• Expanding service delivery Systems Alliance is an indicator of the firm’s deep understanding of the competitive landscape. These differentiating automation assets, it still acts largely siloed
toward AI partners are complemented by exhaustive proprietary tools, in particular HPA in healthcare from the by traditional business units. HfS recognizes that Cognizant
• Use case identification and TriZetto acquisition as well as assets such as BigDecisions AI and SightPrism AI. is working toward a more holistic approach seeking to
matching those with most leverage commonalities between the various platforms
• Deep entrenchment in healthcare: Large scale acquisitions have expanded Cognizant’s deep
relevant tools and approaches. The same applies for building out AI
• Scale and repeatability entrenchment in healthcare. Its HPA platform extends RPA capabilities to AI and machine learning, while
leveraging broader assets, such as vertical-specific BPaaS offerings. HPA offers industrial scale of practices and moving toward a foundational AI layer.
deployments, leveraging the intersection of automation and AI. At the same time Trizetto offers vast While Cognizant is pursuing AI opportunities more
amount of health related data sets that can be use for multiple use cases. High-end use cases include opportunistically, this is limiting the leverage of data sets
diabetic eye screening. Cognizant is leverage deep learning technology, which combines neural network across the organization.
Plan
architectures with massive computing vision technique to allow a software programs to “learn” a • Overemphasis of conversational services: While HfS
pattern from data, delivering a way to automatically identify images of eyes with nodules, lesions, and agrees that AI has the potential to be the next UI, among
Implement
other abnormalities with high accuracy. the various building blocks of AI, Cognizant appears to
• Blending Human Science with Analytics and AI: Cognizant is leveraging human science to establish new overemphasize conversational services at times.
Manage
hypothesis which help data scientists to develop improved and algorithms. What Cognizant is calling Conversational services have a lower complexity then data
Contextual Analytics is going beyond notions of explaining AI as augmented intelligence. This holistic science projects, so a more nuanced communication might
Operate
and cross-discipline approach aligns well with Cognizant’s consultatiuvae approach to Intelligent be more appropriate.
Optimize Automation.
Relevant Acquisitions/Partnerships Key Clients Operations Technology Tools and Platforms
Acquisitions: Cognizant works with clients across industry Geographic footprint and scale of the AI • BigDecisions AI: BigDecisions AI is a data science
• Mirabeau (2016) sectors: practice: platform (Machine Learning as-a-Service) designed
to provide a stable, secure, and high performance
Partnerships: • Top 3 Global Automaker Cognizant’s AI talent spans three main groups: execution environment for deep learning
• AWS • Large US Retailer • Intelligent processes and repeatable applications running on cloud or bare metal
• Google • US Insurance company automation covers approximately 650 FTE infrastructure or Edge devices.
• Microsoft • Global Mining leader AI specialists, with a separate focus on • Sight Prism AI: Cognizant's proprietary Customer
• IBM Watson • US Insurance Company healthcare claims automation through a Journey Analytics Ecosystem, now powered with AI
• Kore AI • Global Financial Services company subsidiary (known as HPA, a subsidiary of advanced analytics capabilities coupled with human
• Artificial Solutions • US Managed Care company TriZetto). sciences, providing, measuring, and monitoring new
• Conversable • US Insurance leader • The Conversational AI group has customer experiences.
• A2iA • European-based Financial Services leader approximately 100 FTE AI specialists. • CuRIE: Customer Retention Intelligence Engine that
• Avaamo • Large US-based Pharmacy • Applied AI has around 3,000+ trained helps identify key customers and optimizes the
• Ayasdi specialists in machine learning, deep contact strategy for improving customer retention.
• BigML learning, NLP, cognitive, and other relevant • Conversational Agent Framework Cspeak: A digital
• CognitiveScale technologies. They are trained in the agent that understands and answers customer
• DataRobot dominant technologies (Cortana, Watson, queries in natural language. It maps the intent of
• Chatterbox Labs etc.) and carry the majority of the AI customer queries with respect to business function
• Clarabridge resources within Cognizant and support and domain.
programs across the company.

29j916d5
Deloitte Global professional services firm with differentiating capabilities around
complex data projects and RPA expansion
Winner’s Circle
• Data management strategies • Consultative approach to evaluating AI projects: Clients feel positive about Deloitte’s consultative • Enhanced geographic consistency: Deloitte’s global
integrating semi and approach that helps to validate both use cases and business cases. This also includes a proactive network of member firms means there may be
unstructured data approach to stakeholder management, demonstrating the benefits of AI and machine learning. opportunities for inconsistency from country to
• Actual delivery of services Deloitte is moving beyond traditional data science to include other fields such as behavioural country. Consequently, delivery capabilities should
• Support with use case science, psychology, and linguistics. Clients also indicated Deloitte’s flexibility to chose the right tools be assessed both on a global as well as local level.
identification and alignment for even limited budgets. Similarly, the quality of delivery might not be
with relevant tools consistent globally.
• Broad recognition for excelling around complex data issues: Clients and stakeholders cite Deloitte’s
• Vision for and investments
in the evolution of IA reputation for having strong capabilities around complex data issues. This is further underpinned by • More nuanced marketing: Probably not least
• Scale and repeatability deep investments into its ClearLIGHT analytics platform and the HexScore AI accelerator platform because of the global network of member firms,
that creates an ecosystem to integrate multiple technologies such as AI, multiple data environments Deloitte’s marketing presence is very high level.
Plan (including open data, sensor data, geospatial data, image date, social media, text), curated data sets, This is in part mitigated, however, by its client
and cloud infrastructure to generate better insights. centricity and technical capabilities.
Implement • Expansion of RPA: Aligned with its consultative approach Deloitte is looking at RPA as an opportunity • More attention to training: Some clients encourage
to help clients begin their cognitive journey. An example for this is its D-ICE solution, which combines Deloitte to pay more attention to training, but they
Manage image recognition scripting, NLP, and RPA to eliminate the risks of data entry errors. also reported this partly mitigated when raised
• Extensive investment program to drive cognitive agents both for clients and internally for legacy with solid transition and knowledge management.
Operate audit and tax businesses: An extensive investment program to leverage cognitive capabilities that
can be extended to virtual agents provides a critical competitive differentiator. These investments
Optimize are used externally to transform clients’ businesses and internally to transform Deloitte’s businesses.
Acquisitions: Deloitte works with clients across Geographic footprint and scale of the AI • Cognitive Advantage: Broad set of capabilities across robotics
• Blab (2017) industry sectors: practice: and cognitive automation, cognitive insights, and cognitive
• Strat (2017) engagement offerings.
• Global automobile manufacturer Deloitte serves clients globally across 150+ • D-ICE: Uses cognitive technology (advanced character
Partnerships include: • Large global manufacturer countries with the following breakdown: recognition, machine learning, natural language processing) to
• Facebook • Top global pharmaceutical • NAM: 61% expedite conversion of unstructured data to a structured
• AWS company • EMEA: 22% format. Used as foundational layer across the various AI
• IPsoft • Health plan provider • APAC: 15% technologies.
• Ayasdi • Global biotech firm • LATAM: 2% • ClearLIGHT: Private, cloud-based, end-to-end analytics
• NarrativeScience • Large global insurer platform designed to give decision makers at every level of the
• WorkFusion • Global airline Deloitte has 4,400+ FTEs focused on organization fast, cost-effective access to actionable
• KnuEdge • Global commercial bank cognitive/AI globally, including 2,600 data information.
• Expero • UK financial services company scientists. Resources are spread across • HexScore: An AI accelerator platform that creates an
• Synthesio • US state government Deloitte’s global offices, along with 26 ecosystem to integrate multiple technologies such as AI,
• Skyhook global delivery centers, 27 greenhouses, multiple data environments (including open data, sensor data,
• SEMOSS and 53 Insight studios and hubs. geospatial data, image date, social media, text) curated data
• Beyond Limits sets, and cloud infrastructure to generate better insights.
• Chatterbox Labs Planned growth rate from FY17 to FY18: • Nora: Virtual agents in-house cognitive solution develops
~30% insights by ingesting data and helps users improve health
outcomes.

29j916d5
Genpact Global professional services firm accelerating AI platform approach
Winner’s Circle
• Actual delivery of services • Mature and holistic approach to IA: Genpact is pushing a holistic approach to Intelligent • Adapt thought leadership to specific vantage points:
• Works with clients to Automation by building out micro-services and machine learning on top of systems of While clients praised Genpact’s proactiveness on
integrate IA into service engagement. This is further enhanced by leveraging NLP and cognitive computing, in innovation, some encourage them helping them
delivery particular for knowledge transformation process situations where Genpact needs to automate imagining the future state of their organizations and
• Vision for and investments context-specific knowledge extraction from unstructured sources of information. Genpact is processes. Thus, Genpact might want to expand their
in the evolution of IA in the early stages of expanding this approach to conversational AI and data science AI. Rage’s automation-centric thought leadership to support clients’
• Scale and repeatability of
computational linguistics assets and automation technology are a critical accelerator on this transformation of knowledge work and beyond.
deployments
journey. • Translating technical requirements into the world of the
• Data management strategies
for semi and unstructured • Strong alignment with OneOffice concept: Genpact is expanding the Lean Digital Framework process owners: While clients are positive on the quality
data with the Cora platform, leveraging an expansive set of AI assets. Those assets include the Rage of service delivery, some would encourage more clarity
and PNMSoft acquisitions. Cora is built in a modular fashion digital core, analytics and AI. about the work done to their processes. However, more
Plan Within AI, Genpact is curating technologies in computational linguistics. They are aligned with generally, they would encourage Genpact to translate
the OneOffice concept by routing customer data all the way to execution. technical requirements into a language that is
Implement • Rage acquisition provides ready to use AI assets: Rage’s focus is not on automating specific understood by the business.
process steps, often on sub-process level, but on developing a broad ranging assets for end to • Balancing the integration of assets like Rage Framework:
Manage end for custom solutions with a deep vertical footprint. Rage is comprising both standalone solutions as well as
• Integration of unstructured data: Genpact has invested heavily in analytics and big data with being core components for Cora. Thus, the culture and
Operate agility are being preserved, but its assets are only slowly
a dedicated research lab in Bangalore, India. They have developed a data engagement
platform using big data technologies in order to be able to harness structured and moving to become a foundational layer across the
Optimize organization.
unstructured data from multiple sources which is also integrated into Genpact Cora.
Acquisitions: Genpact works with clients across Geographic footprint and scale of the AI • Genpact Cora Modular open architecture platform that
• Rage Frameworks (2017) industry sectors: practice integrates automation, analytics and AI engines as well as
• PNMsoft (2015) supporting the Digital products and solutions.
• Endeavour Technologies (2015) • Global wealth management firm Genpact’s overall delivery is 70+ centers • Cora LiveSpread AI tool that speeds up financial spreading for
• OnSource (2017) • Global wealth management firm across 18 countries. They are served by commercial underwriting and risk management by automating
• TandemSeven (2017) • Global financial institution Genpact’s 1,300 data scientists and 1,000 the process each step of the way.
• Global manufacturing firm digital engineers. • Cora LiveWealth AI based product uses patented extraction
Partnerships include: • Global financial institution technology to source information for performance reporting.
• Microsoft • Large CPG firm Locations: Delivery breakdown by FTE • Cora Sequence Workflow management suite automates and
• AWS • Global pharmaceutical company percentage: optimizes business processes streamline and automate
• Salesforce • Large diversified manufacturing APAC (includes Bangalore, Gurgaon, complex processes and transform their business.
• WorkFusion company Hyderabad, Kolkata) – 85% • Cora PharmacoVigilance Touchless AI-based adverse event
• Pega • Large confectionary, food, and Global (includes resources from US, EU) – identification and qualification digital product.
• OutsideIQ beverage company 15% • Intelligent Reporting Assistant Digital product using machine
• SparkBeyond • Large consumer packaged goods learning and data science AI to source and enrich structured
company and unstructured external data.

29j916d5
IBM Leading global service provider driving AI at scale with differentiating assets
around Watson ecosystem
Winner’s Circle
• Vision for and investments • Strong association with cognitive and AI: Through the Watson brand and its capabilities, IBM has • More nuanced and detailed communication of
in the evolution of become synonymous with the advance toward Cognitive and AI. For IBM, Cognitive is a strategic Watson capabilities. Cognitive and Watson are a
Intelligent Automation pillar of its corporate strategy along with cloud. Thus, IBM is broadly being perceived as early central pillar of IBM’s corporate strategy, which at
• Scale and repeatability of mover and thought-leader in the space. times leads to an inflationary use of the cognitive
deployments • Deep investments in Watson capabilities: Deep long-term investments in Watson are further moniker or too generic communication. Some of IBM’s
• Expanding service delivery competitors partnering around Watson capabilities
enhanced by acquisitions including the Weather Channel and AlchemyAPI. The Weather Channel
toward AI articulate their strategy more succinctly than IBM.
• Data management strategies acquisition highlights the investments in data assets that complement technology advances.
integrating semi and • Formalized AI practice: Cognitive Business Decision Support (CBDS) sits within GBS, within this • Watson curtails technology and tool choice: While
unstructured data service line, there is Advanced Analytics, Watson AI & Data Platform, Watson IoT, and SCM, and Watson offers a plethora of capabilities, some clients
Watson Health. This drives AI as a foundational layer across the organization. would like to see a broader choice in tools and
• Driving scale through focus on core technologies: IBM is focusing on three core technologies and technologies. Not least in the context of perceived
Plan
driving them out at scale: Blue Prism in RPA, IPsoft in Autonomics, and Watson as a virtual agent necessary investments for projects. HfS acknowledges
Implement and broader analytics scenarios. Some clients value IBM as a strategic partner that has that there is at times a trade off between technology
demonstrated the robustness of automation without a single outage. choice and robustness of delivery.
Manage • Loss of high-profile Watson accounts: Press coverage
• Broad set of vertical and domain-specific offerings, including Cognitive Network Monitoring for
of the loss of some hospital accounts with significant
telcos, cognitive manufacturing, visual inspection for work quality, and cognitive assignments for
Operate investment levels in Watson, reinforces the necessity
HR.
of a more nuanced communication.
Optimize
Acquisitions include: IBM works with clients across industry Geographic footprint and scale of the AI practice: • IBM Services Platform with Watson:
• AlchemyAPI sectors: Foundational to IBM’s AI-powered automation
• Blekko The AI business (like all of the CBDS Service Line) with single platform integration. Thus, allowing
• Merge Healthcare • Sysco focuses on 14 industries but their most successful to deploy “consumable” services to clients.
• The Weather Company • Danske Bank are financial services, insurance, public sector, Building blocks include IBM Watson and
• Truven Health Analytics • Maersk healthcare, and telco, with most resources in this Analytics as well as specialized Data Lake APIs for
• UStream • Exxon Mobil space located in the US, EU, Japan, and India. handling service management data.
• Promontory Financial Group • KONE • Representative Watson APIs: Include Vsion
• Korean Airlines (visual recognition), Language (natural language
Partnerships include: • Quest Diagnostics classifier, language translator, personality
• IPsoft • MetLife insights, tone analyser, natural language
• SAP • Medtronic understanding), Speech (speech to text, text to
• Light Bend • AmeriSource Bergen speech), Data Insights (retrieve and rank,
• Live Person document conversion), Conversation and
Discovery (enable the user or application to find
relevant information through cognitive and AI
search capabilities.).
• Automation Fabric: Integration between RPA,
BPM, analytics, and cognitive tools.

29j916d5
TCS Leading global service provider demonstrating a broad set of AI use cases
Winner’s Circle
• Vision for and investments • Formalization of Enterprise Intelligent Automation business unit is starting to galvanise TCS’ AI • Investments in marketing: Although clients point to robust
in the evolution of IA efforts as well as driving a much more holistic automation strategy beyond the early focus on delivery and strong technical skills, the narrative on the
• Governance and testing proprietary IP. This is helping to reinforce a holistic approach to Intelligent Automation. The vision broader notion of IA as well as AI capabilities should be
services and focal point is the “Machine First” operating model. While the vision is compelling, it needs to be enhanced. Similarly, TCS should invest in overarching thought
• Actual delivery of services translated into referenceable input that can be driven through the organization. leadership and a broader vision of what IA is meant to
• Scale and repeatability achieve. Furthermore, the relationship between Digitate and
• Integration into client • Clear articulation and strong roadmap for Watson services: A dedicated Watson CoE demonstrates
pre-built PoC/PoT, creates reusable assets, frameworks, and undertakes cognitive research and Enterprise Intelligent Automation remains opaque to
delivery capabilities
Watson advisory work. The TCS Content Lab has also set up a Watson unit for cognitive research with external stakeholders.
a focus on TMT vertical. • More effective guidance and management of stakeholders.
Plan • Broad set of AI use cases: Demonstrates the acceleration of TCS’ AI efforts. The spread of use cases is Clients encourage TCS to be more proactive and assertive
referenced by jointly setting an AI CoE with Intel to assist Nervana AI Academy members, including with its management of stakeholders. They want to be
Implement developers, data scientists, students, professors, and startups in developing AI solutions. Similarly, it inspired by ideas and expect TCS to be more proactive
is helping a client exploring quantum computing techniques on AI algorithms in blockchain beyond delivery. This includes beefing up transformation
Manage applications capabilities, including Design Thinking.
• Focus on industrialization and ease of uses: TCS is working toward a Bot “App” Store and catalogue • Demonstrate platform and tool capabilities. Clients would
Operate for reusable assets. Clients feel TCS is doing a good job in scaling deployments, this includes good like to see more proof points for TCS tool capabilities. They
references on testing cognitive capabilities. encourage TCS to invest more in certification of its people.
Optimize
Partnerships include: TCS works with clients across Geographic footprint and scale of the AI practice: • ignio: Digitate’s Cognitive Automation platform has the
industry sectors: potential of two core capabilities native to the platform. The
• Google • 75+ active customers first is to manage business and IT operations autonomously,
• AWS • American multinational • 1400+ experienced AI professionals and the second is scenario planning with “what if”
• Microsoft corporation • Delivering AI solutions across verticals around the modelling capability to understand impacts on the IT and
• IBM Watson • Large integrated energy globe process environment over time.
• Pega AI company • End-to-end AI servicing offerings including strategy, TCS also has various frameworks and accelerators such as:
• IPsoft • Leading financial corporations CoE setup, platform selection, implementation, and • Cognitive Computing Suite: A set of turn-key tools for
• Kore AI • Leading luxury retailer maintenance provisioning AI-driven business insights. Using cognitive
• Inbenta • British mutual financial • ~40 consulting engagements computing and big data analytics to shorten business
• H2O AI institution • Enabling solutions through cognitive value process lifecycles by automating and simplifying a host of
• API AI • American privately held global assessment activities that currently involve complex decision-making.
• Abbyy corporation • KNADIA (Knowledge driven conversational agent
• WorkFusion • Leading Nordic framework): An Intelligent agent (bots) network integrated
• Stanford NLP telecommunications Beyond enterprise intelligent automation, Digitate is a with a NLP intent layer as well as an interaction layer;
• Core NLP companies TCS venture with ignio as its flagship investment, and is underpinned by a deep learning model, which allows for
• NLTK • Australian telecommunications modelled like a Silicon Valley start-up with more than self-learning capabilities.
and media company 400 employees. Digitate targets Global Fortune 2000 • 360 Degree Assurance: An AI and NLP powered Assurance
enterprises with a dedicated sales force and Automation engine that provides actionable insights and
professional services consultants in each of the target constructive feedback that can improve software testing
geographies. process over a period of testing cycles.

29j916d5
Atos Global service provider accelerating the industrialization of automation and AI
High Performer
• Vision for and investments • Focus on industrialization of service delivery: Atos stands out by driving IA at the heart of its service • Overarching thought leadership: Although Atos’ strength
in the evolution of IA delivery backbone. This includes linking business services and applications orchestration with is industrializing delivery, HfS would encourage to build
• Expanding service delivery infrastructure cloud provisioning. A crucial element in this strategy is integrating big data and operational out narratives for the direction of travel of process
toward AI analytics into the automation approach. innovation. This should be extended to a broader Future
• Use case identification of Work theme leveraging Atos’s strength in infrastructure
• Industrialization of machine and deep learning. While still in an early phase of development, the Bull
matched with relevant tools and, in particular, workplace solutions.
• Scale and repeatability of Ystia cognitive software suite is leveraging Bull HPC capabilities to offer clients a collaborative workspace
deployments that includes the sharing of assets, models, and frameworks as well as pre-trained machine learning. The • Investments in marketing: Atos is not yet seen as a
• Integration of automation strategic intent is to industrialize use cases to enhance the customer experience. The focal point will be significant player in the bigger buying decision on
into delivery capabilities weak signal detection for cognitive data center, security operation center, and video security. Intelligent Automation. Thus, investments in marketing
• Virtual agents as part of a holistic approach to IA: Atos Virtual Assistant (AVA) offering is focusing on are required to convey Atos’ differentiating strategy.
Plan routing customer requests from the client interaction all the way to execution, be it RPA or autonomics. • Demonstrating the proof points: As Atos is pushing the
Currently it is one of the most holistic approaches in the industry that is starting to scale aggressively with boundaries toward less well-developed and therefore
Implement key customers. less-documented use cases such as application
• Knowledge integration at the core of Enterprise AI: Atos is aiming to move from system integrator to management, operational analytics, and virtual agents,
Manage
knowledge integrator. Adopting knowledge graph technology is the lever to advance to industrial scale the company needs to demonstrate proof points, metrics,
knowledge integration and machine reasoning. Over time those knowledge graphs will train machine and insights from the early deployments.
Operate
learning models and, conversely, machine learning will help with knowledge graph construction
Optimize
Acquisitions include: Atos works with clients Geographic footprint and scale of the AI practice: • Bull Ystia: “Is a fast track to AI with multi-use case cognitive software
• Convergence Creators Holding across industry sectors: suite, leveraging machine learning and deep learning to enable AI on
(CVC) (2018) Main locations are India (Pune, Bangalore), UK multiple environments. Positioned as the cognitive software suite in
• zData (2017) • Ulsterbank (London), France (Grenoble), Netherlands the Codex portfolio. Includes Studio, a cognitive application
• Multinational banking (Groningen). The delivery of AI/RPA is via Atos’ division development self-service; and Forge, a common workplace to store,
Partnerships include: and financial services delivery channel: share, retrieve, and update assets, models, frameworks. The strategic
• IPsoft holding intent is the industrialization of use cases.”
• Arago • Finnish multinational Global geographic scope with operations in 72 • Atos Virtual Assistant (AVA): Delivers machine learning capabilities
• Google communication and countries. in the Cloud and enhances the end-user experience, for instance by
• Microsoft Information technology Delivery arranged through global business units (GBU) automatically answering IT service desk and data center requests
• AWS as follows: through a virtual agent. Leverage of CogniCor and Living Actor
company
• Salesforce Einstein • UK and Ireland technologies.
• Retail post office
• LivingActor • Iberia • Atos Intelligent Automation Platform (AIAP): Atos is extending its
• CogniCor
company • France internal delivery model for managed services into a customer facing
• Pega • A municipality in the NL • North America platform. The emphasis is on standardizing on ServiceNow and these
• thoughtonomy • An executive agency of a • South America orchestration capabilities are directly linked to autonomics solutions
• Celaton ministerial department • Benelux from Arago and IPsoft including piloting virtual agent technologies.
• Expert System • German multinational • Central and Eastern Europe For AIAP these capabilities are expanded by expansive cognitive
• Sinequa pharmaceutical and life • Germany capabilities as well as RPA.
• OntoText sciences company • Asia Pacific (including Australia)
• India, Middle East, and Africa

29j916d5
Capgemini Global service provider accelerating the industrialization of Intelligent
High Performer Automation and AI
• Tool and platform strategy • Focus on industrialization of automation and AI: Capgemini is accelerating the industrialization of IA • Siloed organizational structure: While working toward
• Data management strategies and AI. In particular, the Automation Drive framework is helping to build and leverage commonality holistic automation, the organizational structure is still
supporting integration of across processes. By cutting across the traditional organizational stovepipes, Capgemini has a strong deeply siloed. Automation Drive is starting to address some
• semi and unstructured data alignment with the OneOffice concept. A Group Automation CoE catalogs and proliferates the of those issues, but it is work in progress. Consequently,
• Expanding service delivery automation services across SBUs, but the historical organizational complexity remains. some clients struggle with the organizational complexity.
toward AI
• Integration of Intelligent • Strong emphasis knowledge transfer: Capgemini views knowledge as the critical element for • Overarching thought leadership, vision: While Capgemini
Automation into clients’ institutionalizing AI. The partnership with Transversal links NLP with an expansive knowledge base. is accelerating its journey toward a holistic automation
delivery capabilities • Differentiating focus around Watson for manufacturing and IoT: A strength is its progress on approach, it should expand its thought leadership and
• Scale and repeatability of cognitive IoT solutions in digital manufacturing, connected products, and asset management, true to narrative that helps clients envision the innovation journey
deployments its historic strengths in manufacturing, CPG, supply chain, and logistics. Capgemini is working as and future state, in particular demonstrating use cases and
release partner for IBM’s new Cognitive API services in Digital Manufacturing. Capgemini has strong lessons learned from the early deployments. Furthermore,
experience with Watson Explorer with examples of using its AIE organization to drive Design Thinking clients encourage Capgemini to provide a vision beyond
Plan technology, specifically including dealing with unstructured
workshops to help clients identify and prototype solutions.
• Network of Applied Innovation Exchanges (AIE) is fostering co-innovation with clients. By blending data.
Implement
client experience workshops and Design Thinking, Over 60% of requests for AIE services lead to • Demonstrate AI capabilities locally: While clients are
Manage solutions containing an AI elements. Capgemini’s 5 senses framework is proving a helpful way of positive about the technical capabilities of Capgemini, they
framing those discussions. would like to get more access to Capgemini’s staff on a
Operate proximity basis. Hiring relevant talent locally should be a
priority for Capgemini.
Optimize
Partnerships include: Capgemini works with clients across Geographic footprint and scale of the AI practice: • Automation Drive: Capgemini is blending a broad set of
industry sectors: AI capabilities into the Automation Drive (AD)
• IBM Watson • Over 2,700 staff involved in artificial intelligence framework. The formalization is accelerated through AD
• Microsoft • European Family Pensions Govt delivery from Capgemini’s global community of Store, which is providing a global source for tools and
• Google Agency over 6,300 intelligent automation resources. solutions; key goals for the AD Store are a rapid
• AWS • Nib NZ Health Insurance • Global coverage with the bulk of capability deployment and standardization. AD Library is offering a
• Kore AI • MGEN supporting US and European clients whilst having though leadership hub for +1,800 use cases enhanced by
• Loop AI • Oil & Gas Major steady growth in LatAm, Canada, and Asia Pacific. a virtual agent. AD Academy is fostering training leading
• Senseforth • Automotive Manufacturer • The development of intelligent automation to the development of tailored AI solutions. Together,
• Cortex • Allianz (including AI) is driven top down and governed by these pillars govern and form the foundation of
• Automic • Unit 4 Capgemini Group under the Automation Drive Capgemini’s automation approach across the entire
• Celaton • Ethias initiative. Group.
• Ayehu • Interim services company • Intelligent automation innovation is driven bottom • SmartQA: Self-learning test ecosystem with
• ThinkIQ • Cruise line up by business units designing and building cognitive/analytics capabilities
• Diota solutions based on their respective domain • Intelligent Test Automation Platform (ITAP)/ Autohub:
• KNIME expertise coupled with our globally consistent Test automation platform for DevOps.
approach to innovation that flows from our
Applied Innovation Exchange.

29j916d5
DXC Technology Global service provider accelerating the industrialization
of AI with strong focus on analytics
High Performer
• Actual delivery of services • Driving factory approach to get to scale: At the heart of DXC’s Intelligent Automation approach • Catching up on Intelligent Automation: Having been a pioneer on
• Governance and testing is a factory approach to scale out automation and AI assets. The newly announced Bionix and RPA, DXC needs to further accelerate investments in service
services Agile Process Automation offerings aim to push AI as a utility and as utility services. Focal points delivery but more importantly the demonstration of proof points
• Addressing training and are configurable solutions, service orchestration, and API-based integration. An example for for successfully deployed Intelligent Automation and AI projects.
supporting clients that is the partnership with Microsoft to integrate the Microsoft Cortana Intelligence Suite and
• Works with clients to • Driving automation and AI across organizational boundaries:
Azure infrastructure services. This has improved DXC’s AI model orchestration capabilities, data While DXC clearly is accelerating its automation and AI efforts,
integrate IA into their
infrastructure, and compute and storage capacity. much is driven by the traditional business units. In order to move
delivery capabilities
• Leveraging strength in analytics: Given the well documented issues around the merger of HPE toward the OneOffice, collapsing organizational boundaries is
and CSC, DXC is playing catch up with the leading providers on Intelligent Automation and AI. critical. The Bionix program is meant to change this.
Thus, it is playing to its strength by strongly emphasizing its analytics capabilities predominantly • Driving AI beyond sweet spot in analytics: Having DXC’s analytics
Plan from the old HPE side. The majority of those services are industrialized and commodity services. team driving most of the AI efforts (that is Industrialized AI)
However, DXC has also demonstrated some strong delivery capabilities for high-end data reinforces the necessity to advance toward a cross-organizational
Implement science projects in the health sector. AI strategy. Autonomics and RPA extension are the obvious
• Blending analytics and AI into vertical offerings: What DXC is calling “Applied Industrialized AI” targets. Again, Bionix will address some of those issues.
Manage is aiming to leverage the expansive analytics assets to build out verticalized solutions. • Demonstrating the willingness and ability to co-innovate with
• Leveraging information governance experience into broader AI governance: Evolving from its clients: Unsurprisingly, the merger had seen some clients left with
Operate sweet spot in analytics, DXC has a broad set of capabilities around the broader notion of delays and lack of clarity for joint investment decisions.
information governance.
Optimize
Partnerships include: DXC Technology works with clients Geographic footprint and scale of the AI • DXC Bionix: While being work in progress, Bionix blends data
across industry sectors: practice: science and lean methods across IT operations and BPS. Focal
• Microsoft points are data mining to drive discovery processes and
• AWS • Large European hospital Currently DXC has around 750 delivery monitoring. Over time, the offering will evolve to chatbot
• IBM • Large state welfare agency resources driving the Bionix efforts. integration and data orchestration. RPA and intelligent bots are
• Hitachi Vantara • Large bank in Asia being leveraged to automatically diagnose and resolve incidents.
• Alterix Furthermore, DXC has around 8,000 Bionix is the single biggest internal investment within DXC.
• H2O.io analytics resources globally. Analytics and AI • DXC Industrialized AI: Industrialized AI allows the enterprise to
• Anaconda are embedded in most solutions that we learn from data, improve business decisions, and automate
• Datasift bring to the market and it is difficult to complex operations. DXC Industrialized AI provides a strategy
• Digital Barriers completely isolate AI within that. Of those, workshop that identifies high-value data stories; customized
300-350 are purely driving the development technical infrastructure, designed to speed the development and
of AI solutions led horizontally by Analytics deployment of enterprise-scale experiments, curated sources of
Practice, which oversees development in data, fully-functioning AI applications, and consulting services that
delivery engineering and industry/offering maintain the solution and transform the organization.
verticals plus approximately 500 client • DXC Analytics Platform: A fully integrated, industrial-strength
delivery resources. analytics and artificial intelligence (AI) platform that includes
enterprise-grade architecture, multiple deployment options, end-
to-end managed services, and guided onboarding.

29j916d5
EY Global professional services firm with differentiating capabilities around
High Performer domain-specific projects and RPA extension
• Leverage of Design Thinking • Focus on strategic bets: EY is still early on its journey toward AI, but is leveraging its mature client relationships and • Lack of formalized approach to AI: EY is
for reimagining and selective bets on domain-specific capabilities as well as the depth of its relationship with Microsoft. It has a strong lagging its peers in driving a formalized
transforming processes focus on financial services and differentiating capabilities around manufacturing. In the discussions with HfS, EY stood approach to AI through the organization.
• Governance and testing out by a realistic depiction of client discussions and requirements. Broadly speaking, it is expanding its analytics While the AI leadership team is
services capabilities to drive knowledge discovery, that is finding new answers to complex problems. Clients are positive about demonstrating a succinct vision, it needs to
• Training and supporting the delivery capabilities. demonstrate its ambitions to the broader
clients
• Strong commitment to Microsoft: Building on the focus on mega ISVs from it analytics business, EY has a close market.
• Actual delivery of services
collaboration with Microsoft, actively working with them to deploy Cortana as well as developing standard automation • Alignment of data assets: Given the relative
for document reading (invoices, checks, etc.). organizational immaturity, EY has not yet
• Extension of RPA: EY has developed a strong reputation for driving RPA deployments with a transformational and end- fully decided how to align data assets with
Plan to-end point of view. It has an alignment with the OneOffice concept by its partnership with TrustPortal and by AI capabilities. However, it is leaning
leveraging a broad set of cognitive tools such as ABBYY and Arago’s Hiro for parsing, Workfusion and Azure Machine toward building out domain-specific assets.
Implement Learning Studio for machine learning, and big beasts like Watson for a broad set of services including virtual agents • Articulating and demonstrating the
and chatbots. internal change management: EY is
Manage • Blending human centred on ethically aligned design: Utilizing its Wavespace Design Thinking centers, EY drives investing significantly to transform its tax
ethically aligned design. It is also in the steering committee of an IEEE working group around the topic. An example of and advisory business through AI, but it
Operate the application is securing trust in underwriting delivered through broad Watson capabilities. needs to demonstrate the ambition and
• Expansive push on AI training: The EY AI academy project is driving training at scale through the organization. Badges proof points.
Optimize scheme and mapping of trained employees to Data Science and AI base capabilities and offerings.
Acquisitions: EY works with clients across industry Geographic footprint and scale of the AI • EY Synapse: EY Synapse, their global Analytics-as-a-Service (AaaS)
• Bluestone (2015) sectors: practice: platform, accelerates speed to insights by operationalizing data
• C3 (2015) and advanced analytics quickly and cost effectively on an
• CogniStreamer (2017) • American gas company AI practice comprises nearly 3,000 practitioners Analytics-as-a-Service platform. EY Synapse combines third party
• Multinational oil and gas company (33% in the Americas, 51% in EMEIA, 14% in and proprietary technologies, EY-developed IP assets, advanced
Partnerships include: • American Asia-Pacific, and 2% in Japan), a literal “team of analytic approaches, and deep sector knowledge. By adding EY
• Microsoft pharmaceutical company teams” in cities around the world, from AI to capabilities around the behavioral aspects of data and advanced
• IBM Watson • Biotechnology company Intelligent Automation. EY takes an integrated analytics, EY aims to provide scope, scale, and support to business
• SAP • Global cruise company approach that focuses more on client needs leaders.
• GE • Mining and metals company and outcomes than internal practice lines and • Smart Information Retrieval Engine (SIRE): The technological
• Pega • European energy company boxes. system to automatically extract data fields from unstructured
• European bank documents such as scanned contracts. Packaged solution usable
• English financial services company Offerings are co-developed, co-managed and through a web-based interface and hosted on EY secured
• Multinational chemical company co-delivered with resources from across digital environment and allows for a complete audit trail. Based on
• National law enforcement agency technology, customer experience design, RPA, Image cleansing and pre-processing, optical character
• Global manufacturing and services AI, corporate strategy and risk, in tandem with recognition, document and layout analysis, and natural language
conglomerate functional and sector experts, effectively processing.
• National public transportation creating a global intelligent automation • EY OPTIX Internal Audit analytics platform: EY Optix facilitates the
authority network of excellence that brings together ability to integrate standardized foundational analytics-based testing
skills, IP, infrastructure, assets, capabilities, and into core business process audits. It uses machine learning to cluster
ecosystem partners. customers in relation to the risk of having a payment delay.

29j916d5
HCL Global service provider accelerating industrialization through an
expansive set of automation and AI capabilities
High Performer
• Tool and platform strategy • Broad set of automation and AI capabilities: HCL’s Intelligent Automation and AI capabilities are among • Move beyond tool and technology mindset: While HCL
for Intelligent Automation the broadest in the industry ranging from RPA to autonomics to test automation. The challenge is to has an enviable set of technology capabilities, HfS
• Vision for and investments evolve these capabilities into a narrative resonant with process owners with specific use cases. HCL’s recommends to drive stakeholder discussions more
in the evolution of ambition is to evolve DryICE toward scalable deployments that can be consumed in a plug-and-play style around business problems and outcomes. Some clients
Intelligent Automation model. The complexity is indicated by 45+ technology components encompassing IA such as cognitive suggested HCL has a technology centric communication
• Leveraging service computing, predictive analytics, deep search, and NLP. The capabilities even include a physical robot and should provide more clarity on commercial issues.
orchestration to deliver end- (Melvin). • Simplifying the nomenclature: Naming and product
to-end and point solutions
• Governance and testing • Flexibility to provide end-to-end and point solutions: HCL has a strong emphasis on orchestration and marketing need a refresh. The array of acronyms and
services integration of automation capabilities; this can be particularly seen on two levels: 45+ modular blocks of products can be confusing at times. In particular, clarity
microservices/microapps that are the key lever to become “silo agnostic,” the Gold Blue Print (GBP) is on DryIce has been called out by clients. Is it a platform, a
Plan the process ecosystem of HCL Best Practice processes and the Service Xchange to drive broad service framework, or has it evolved toward an umbrella
orchestration. HCL ServiceXchange uses SnapLogic iPaaS, SX Integration Hub, and IBM Cognos Analytics branding?
Implement to help CIOs achieve expansive service orchestration. Some clients praise HCL for adapting contract • Scalability: While acknowledging the wealth of
terms even if it means cannibalization of revenue streams. technology assets, some clients encourage HCL to
Manage • Blend of consultative approach, talent, and technology assets: Clients were positive about the demonstrate the scalability of its approach, especially
consultative approach in many engagements. In particular providing the right talent guiding clients outside its sweet spot in infrastructure. This ties in with a
Operate through the technology implications as well as the available assets. more outcome focused communication.
Optimize
Strategic investments include: HCL works with clients across industry sectors: Geographic footprint and scale of the AI • HCL DryIce framework: Proprietary framework that
• Alpha Insight practice: leverages a reference architecture of process
• Moogsoft • Statoil automation and process transformation.
• AIB HCL suggests 54% existing customer penetration, • COPA: Is an end-to-end process automation
Partnerships include: • DNB 90% of all new implementation and maintenance platform utilizing Lucy, HCL’s cognitive assistant
• AWS • Deutsche Bank undertaken. solution developed on IBM Watson’s
• Google • Volvo Group conversational services to deduce voice-based
• Microsoft • NXP Core R&D development on AI, machine learning, inputs to work with different legacy systems and
• IBM Watson • Estee Lauder deep learning, algorithm development, applications while also identifying patterns and
• WorkFusion • UPM computer vision and NLP with leading high tech, making sense of the unstructured data.
• Haystac • Novartis ISV, and FinTech companies. • AIOPS: AIOPS is AI driven technology operations
• AppDynamics • Cummins that uses smart algorithms to provide zero touch
• Dynatrace • American consumer goods corporation • $50M+ investments in engineering labs automation of tasks and processes.
• Redlamda headquartered in USA • 22+ labs across geographies • iAUTOMATE: A standalone “core A.I.” module for
• Moogsoft • American medical technology company that • 100+ development centers with dedicated smart automation for issue resolution, automated
• Dynatrace manufactures and sells medical devices and labs response etc.—powered by an extensive
instrument systems • 3,000+ data scientists delivering advanced knowledge and script base.
• A Japanese technology company and contextual analytics • Lucy: Cognitive interface leveraging IBM Watson,
manufacturing business and industrial Amazon Alexa and other standard tools to provide a
imaging products human-machine interface transformation framework.

29j916d5
Infosys Global service provider with differentiating capabilities in data
High Performer management and analytics
• Expanding service delivery • Excellence in data management and analytics: Infosys has a strong reputation for its technical capabilities • Infosys’ corporate challenges led to management
toward AI among its clients. Some go as far as describing Nia “as almost a complete solution” that allows clients to distractions and delay of investment commitments.
• Vision for and investments build strong models. However, clients are equally suggesting those capabilities come with a hefty price tag. Some clients report about delays in decision making
in the evolution of • New leadership team is rebalancing automation strategy: The new automation leadership team installed and investment commitments. Infosys urgently needs a
Intelligent Automation after Vishal’s departure is rebalancing Infosys automation and AI strategy and driving the company more phase of organizational stability.
• Data management strategies
into the mainstream. The main changes that have been identified as priorities include: 1. Make Nia more • Strategic pillars of Automation and AI need to be
supporting semi and
unstructured data modular to address concerns on pricing, 2.Provide a more effective integration of Nia and AssistEdge and explained in a more tangible and meaningful way.
• Governance and testing 3. Become more open to third-party tools. In particular the last issue will move Infosys more into the Given the importance for Infosys overall strategy and
services mainstream and align the company with buyers’ expectations. the prominence in external communications, the lack
• Evolving toward an automation ecosystem: The new leadership team is building on the strength of Nia of clear narratives remains surprising. The emphasis
Plan but is expanding it toward a broader partner ecosystem. Microsoft and IBM Watson are the focal points of should be more on business problems and outcomes
these efforts. The notion of the ecosystem also works the other way around as Infosys has licensed Nia to rather than technology. HfS recognizes that the new
Implement one of its competitors. management running the automation strategy is
• Driving a holistic approach with data curation as starting point not by-product: Nia for Knowledge - starting to remedy this.
Manage enables enterprises to capture, formalize and process knowledge and its representation in a ontology • NIA is seen as expensive platform by some clients.
based structure. This allows for a holistic approach to IA, including the ability to broadly scale process Especially where discussions start around RPA with
Operate mapping. Data curation is the starting point and centrepiece of Infosys’s IA strategy. This is amplified by a much lower investment levels, some clients would like
machine learning and data scientist workbench. to see more effective stakeholder management and
Optimize demonstration of the value delivered.
Acquisitions include: Infosys works with clients across industry Geographic footprint and scale of the Artificial • Nia platform consists of the following 5 assets:
• SkyTree sectors: Intelligence practice • First generation AI platform – An integrated
artificial intelligence platform incorporating big
Partnerships include: • Leading sports association • Over 5,500 AI professionals, working on different AI data/analytics, machine learning, knowledge
• Microsoft • Global chocolate manufacturer platforms and services across all geographies with management, and cognitive automation.
• IBM Watson • Leading biotech company majority in US, India, and Europe, are delivering • AssistEdge: Provides end-to-end RPA. Uses
• UNSILO • Global financial services company customer projects in the space of AI and integrated software robots to automate any high-
• H2O • Multinational conglomerate automation. touch, repetitive processes.
• Trifacta • Global healthcare company • AI development is facilitated by following a hub • DEEP—(Data Extraction and Enhancement
• Waterline Data Science • Multinational semiconductor company Platform): A platform that ingests heterogeneous
and spoke model. A central organization drives AI
• TidalScale • Global telecom major source documents or images containing structured
• Wanda • services for clients as well as for internal purpose.
Leading automotive supplier and unstructured information.
• SAS • Global telecom major AI is further democratized along with rest of the • IIMSS—(Infosys Infrastructure Management
• Global specialty chemicals company traditional services at Infosys. Each service line has Services): A unified IT operations command center
built AI-specific Center of Excellences (comprising for data center, infrastructure, cloud, applications,
~30 evangelists) that not only provide external security, network, and business services.
customer services but also drive usage of AI for • SkyTree: Advanced high-performance machine
internal efficiency gains. learning with automated selection of algorithms
and methods to achieve best possible predictive
accuracy.

29j916d5
KPMG Global professional services firm accelerating formalized AI practice
High Performer
• Training and supporting • Accelerated organizational maturity with a formalized portfolio of AI practices, tools, technologies, • Maintaining high value approach in a nascent
clients accelerators, and talent globally: By launching KPMG Ignite, the company is demonstrating its market: While KPMG’s consultative approach and
• Data management strategies investments as well as a mature organizational approach. This is enhanced by a broad set of analytics focus on data science projects is strategically sound, it
supporting semi and capabilities including the Sensory Advantage Ecosystem and the Intelligent Signal Engine. has to navigate a nascent market that is awash with
unstructured data • Scale of deployments: Some of KPMG’s deployments demonstrate significant scale of data and PoCs for low level conversational services and RPA
• Actual delivery of services extension. Balancing short and long term implications
complexity. For instance, for a global bank it is driving a centralized data approach by integrating a
• Expanding service delivery profitably will be challenging.
Hadoop cluster to feed into every customer channel. This is reinforcing differentiation about complex
toward AI
data requirements. • Audit might curtail best-of-breed partnerships:
• Scaling out of Watson capabilities: Building use cases around its advisory, tax, and audit business by its Almost unavoidable, any success on the audit side of
Cognitive Automation Lab in New York. Additionally, KPMG’s Astrus is a Watson powered capability to the house might curtail the ability to work with best of
Plan deliver risk assessment services to clients. breed providers. A broad partner ecosystem is
• Effective consultative approach that supports stakeholder management: Clients are positive about how buffering some of these concerns.
Implement KPMG is helping them to drive stakeholder management. This includes helping with governance and • Expand from advisory to broader delivery: Although
setting up a steering committee. Furthermore, it helps with a communication plan as well as anticipated KPMG has made progress in the as-a-service market,
Manage stakeholder push back. HfS would encourage a scaling out of robust
• Extensive investment program to drive cognitive agents both for clients and internally for legacy audit industrialized offerings. The early investments and
Operate experiences in cognitive computing lend themselves
and tax businesses: An extensive investment program to leverage cognitive capabilities that can be
to business process–centric scenarios and use cases.
extended to virtual agents provides a critical competitive differentiator. These investments are used
Optimize externally to transform clients’ businesses and internally to transform KPMG’s businesses.
Acquisitions: KPMG works with clients across Geographic footprint and scale of the • Ignite: KPMG’s portfolio of artificial intelligence capabilities. This portfolio
• Cynergy industry sectors: AI practice: centers around the Ignite platform which brings together KPMG built
• Link Analytics components and accelerators and leverages a deep ecosystem of technology
• Large US financial services Global AI community of approximately partners. The platform is enabled by people skilled in data science, software
Partnerships include: organization 13,000 people with broad capabilities engineering, cognitive technologies, and domain expertise who can utilize
• Microsoft • US utility provider to support these projects around the the platform to deliver client solutions solving specific business problems.
• Google • Large US retail bank world. • KPMG-built accelerators: Patterns and tools that enable rapid AI solution
• IBM Watson • Global online payment development and delivery. These accelerators reduce implementation time
• Oracle processing company 1,300 of those work for the KPMG by integrating with existing IT infrastructure without the need for developing
• Appian • Large US city Lighthouse. The Lighthouse is a Centre new methodologies and templates.
• IPsoft • Major US sports league of Excellence that serves as an internal • A set of frameworks and methods that describe and guide how we approach
• Workfusion • Global insurance company hub between KPMG’s member firms client-specific AI solutions and make them repeatable.
• Nuix • Global biopharmaceutical around the world. It provides • Highly skilled resources, including a dedicated Intelligent Automation Lab
company specialised talent of consultants, within KPMG Ignition, that utilize best tools and approaches to build AI
• Global financial services firm software engineers, data engineers, solutions.
• Global hypermarket chain data scientists, and visualization • Ongoing testing, prototype development and innovation on emerging AI
specialists. tools and approaches.
Cognitive Labs are staffed with specialists • Intelligent Automation Workbench (IAW): Captures the frequently
in AI, cognitive technology, performed cognitive tasks by cognitive scientists and subject matter experts
computational linguistics, and natural in a set of Cognitive Automation Patterns that are encoded in a productivity
language processing. enhancing workbench.

29j916d5
PwC Global professional service firm driving data-centric projects in selective
verticals at scale
High Performer
• Addresses training and • Robust consultative approach: PwC is effectively leveraging its deep client relationships. Clients appreciate • Playing catch up in Intelligent Automation: PwC is
supporting clients that while PwC is providing them with a sound vision on AI, they are not straight away looking to lagging most of its peers in terms of having built out
• Data management strategies aggressively sell to them. Similarly, they value that consultants constantly challenge and question premises broader Intelligent Automation capabilities. While
supporting the integration of as well as taking clients through Design Thinking workshops and other methodology to re-imagine solid on advisory work, we don’t see the same level of
semi and unstructured data processes. Clients are also positive about the proactive way in which PwC is helping them to evaluate investments and commitment as yet compared to its
• Governance and testing relevant use cases for AI. Big Four peers.
services
• Expanding analytics with strong vertical approach: PwC has built out a solid set of AI use cases by • Moving beyond highly domain specific services:
• Leverage of Design Thinking
and co-ideation expanding its analytical capabilities with a strong sector focus. Clients are positive about the scale that While PwC is rightly focusing on a people-centric
PwC is reaching in those use cases. Some of those more complex use cases include NLP for cancer research approach, its limited investments in conversational
and using multi-agent systems to simulate 200K+ go-to-market strategies for autonomous vehicles.. services and industrialized services curtails its ability to
Plan • Recognized for complex data scientist centric projects: Clients appreciate the quality of the work PwC is expand its footprint as the industry is awash with
delivering for complex data centric projects. This includes the technical knowledge around statistical PoCs. The same applies for expanding RPA projects.
Implement models and other AI capabilities such as deep learning, natural language processing, and multi-agent While this approach reinforces its high value approach,
systems, and running these at scale on large datasets. Clients also called out PwC’s willingness to engage the lack of broader investments means that PwC has
Manage with disparate local teams to help with global rollouts. to focus on its niches in its selected verticals.
• Strong internal leverage for assurance, tax, forensics, and advisory. While PwC has built an expansive set • Productizing projects: PwC is still early on its journey
Operate of solutions for those areas, the committed investments appear to be more limited than we have seen to productize projects. It is lagging its peers on
with its Big Four peers. repeatability and broader leverage of assets.
Optimize
Partnerships include: PwC works with clients across industry Geographic footprint and scale of the AI • PwC Apps Marketplace: Launched internally in 10+ territories
• Microsoft sectors: practice: with a portfolio of 100+ global apps and 40 app platforms that
• Google have been identified for development. These apps support repeat
• Narrative Science • Fortune 500 auto manufacturer PwC has more than 5,300 practitioners uses of AI and other advanced applications for modular, scalable,
• Arria • Global pharmaceuticals company globally with skills to address client D&A and and repeatable solutions. Key investments are being deployed in
• Automat • Leading entertainment provider AI ecosystem needs including platform platforms that integrate and accelerate apps within the platform,
• Cortical.io • Global fortune 500 insurance strategy, design, prototyping, algorithmic such as customer Insight, and SCOOP (Supply Chain).
• Agolo provider tuning, deployment, management, • Insights Platform: Analytical apps that can be turned into
• Claralabs • Multinational IT firm architecture, change management, managed services for repeated use while PwC hosts the data,
• DataRobot • Multinational energy and technology operations, network, cloud, security computation, and analysis related to AI.
• AnyLogic conglomerate services, and data analytics. • Data Services: Proprietary data sets created on a combination of
• National health insurance company public and proprietary data used in combination with PwC’s agile,
• S&P 400 life insurance company Approximately 3,000+ staff are working in global data services.
• Multinational beverage corporation analytics (of which ~20% have core AI skills). • AVA Platform: A suite of AI modules designed to automate the
• Global financial services firm This core AI talent pool is projected to reach analysis of tax filings to improve productivity of the tax team.
35%-40% by the end FY18. • Datasieve Platform: A toolkit built with continuous learning to
automate contract review for PwC Assurance.
Artificial Intelligence Accelerator (AIA): • Deals and Disruption Radar: A suite of deep learning models
research center consisting of teams of data tracking structured and unstructured data to monitor activities
scientists piloting solutions using emerging for start ups, VCs, and established companies.
AI techniques.

29j916d5
Syntel Challenger organization driving industrialization with its flagship
SyntBots platform
High Performer
• Tool and platform strategy • Pioneer with holistic automation platform: Syntel’s SyntBots platform is embedded in their managed • Low brand equity and lack of strategy refresh by Syntel:
• Governance and testing services strategy and drives broader digital transformation for clients. SyntBots stand out as an early Although Syntel is willing to be a challenger, it should
services example of a holistic IA platform. However, it has to demonstrate its expanding capabilities as the invest in initiatives to boost its brand equity to get
• Scale and repeatability of market is building out assets at an astounding pace. broader access to stakeholder discussions. Similarly, the
deployments • Alignment with intelligent automation continuum: Syntel’s SyntBots automation platform is well change in CEO has not yet led to a strategy refresh.
• Works with clients to
aligned with HfS’s notion of Intelligent Automation and thus provides an example of a holistic approach • SyntBots needs refresh: SyntBots was a lighthouse
integrate Intelligent
to process automation, from simple scripting up to automated DevOps and autonomics. The fact that investment at inception, but the market is changing.
Automation into delivery
capabilities SyntBots leverages a library of industry-specific business cases, test data, business rules, and other SyntBots needs to adapt, specifically in the context of AI.
components means it was designed to be inherently scalable, with these reusable libraries coming to In our view, plug and play will only address some of a
bear at different points within the client enterprise. At the same time a broad set of microservices client’s requirements.
Plan including Triage and Remediation Assistants and Interactive Knowledge Central are expanding service • Clients require consultative approach: Including Design
delivery capabilities toward AI. Thinking to reimagine processes and not just technical
Implement • Willingness to be a challenger: Syntel is willing to commit to outcome-based SLAs, often domain led. capabilities. Thus, Syntel needs to emphasize its
Thus, Syntel is willing to take on a challenger position, taking advantage of its comparatively low consulting capabilities to complement its technology
Manage installed base. This includes committing to cost savings up front, for instance, where clients had solutions.
maximized efficiencies and savings from labor arbitrage–based models. • Expanding toward project-centric work: Syntel should
Operate • Deep entrenchment with marquee clients: With marquee clients, Syntel has achieved long-running, expand its managed services capabilities toward more
deep engagements in IA. This references the robustness and scale of the SyntBots platform. The project-centric service to support client more broadly.
Optimize
strongest traction for SyntBots is on the DevOps side.
Syntel has a broad set of Syntel works with clients across Geographic footprint and scale of the AI practice: • AI Workbench: User-friendly integration of AI services
partnerships, including: industry sectors: and relevant APIs using out of box UI enables users to
Syntel's Service Transformation Group drives the design, train, and deploy bots on any data set. 60+
• AWS • American Express intelligent automation initiatives with NLP, machine algorithms are designed and optimized to handle
• CA • State Street Corporation learning, vision, and autonomics across different dynamic use-cases from various industries. Pre-
• Google • FedEx industry domains like banking and financial services, processing is one of the pain area in unstructured text
• Hortonworks • Fortune 100 health insurer insurance, healthcare and life sciences, retail, logistics, and scanned documents, which is completely handled
• IBM • The world's leading global transfer and travel. This team focuses on development of AI- by workbench automatically.
• Informatica agency • Enterprise Knowledge—Data Lake: Centralized
enabled automation and domain solutions through the
• Microsoft • Fortune 100 insurer Knowledge Repository and integrated with rule engines
• • SyntBots platform. The team is spread across global
Microstrategy Leading US healthcare provider and algorithms for dynamic servicing. Chatbots’ general
• Oracle • Leading global independent credit locations to scale up automation initiatives across our intelligence and domain data is part of a knowledge-
• Pega rating agency client base. The core SyntBots platform development lake, powered with specific search and parsing
• Pivotal • Leading US truck manufacturer team consists of about 300 people across automation, AI algorithms.
• PTC • Fortune 50 retailer architecture, data scientists, industry domain practices, • Enterprise Integration and Security: All AI capabilities
• Salesforce IT service practices. are delivered as services that can plug and play with
• SAP existing enterprise investments. The platform has been
• Tableau reviewed for security, scalability, and compliance to
• Talend enable accelerated enterprise adoption.
• TIBCO

29j916d5
Tech Mahindra Global service provider with expansive set of AI capabilities and
differentiating focus on TMT vertical
High Performer
• Tool and platform strategy • Strength in industrialized services and assets: Tech Mahindra has a broad set of delivered • Overarching thought leadership and vision: As Tech
for Intelligent Automation projects across industrialized services including monitoring, autonomics, and chatbots. Mahindra is accelerating its investments in automation
• Governance and testing TechMahindra’s strategic view is that AI is the key lever to enable automation. and AI, it should build out an overarching thought
services • Set up for scale: The broad IP, but also a change in the balanced scorecard for its executives leadership depicting the direction of travel for clients,
• Leveraging service stipulating percentages of automation projects, sets up Tech Mahindra well for scale although its as well as the transformation of knowledge work.
orchestration to deliver end- • Getting access to stakeholders: Stakeholders tend to
projects are still in ramp-up mode.
to-end and point solutions
• Scale and repeatability of • Deep focus in TMT: TMT is not only Tech Mahindra’s heritage in but the company has built out gravitate to providers with an earlier and bigger
deployments an expansive set of vertically relevant capabilities. Thus, TechMahindra is focusing on supporting visibility. Thus, a combination of marketing and
the “network of the future” which provides a clear differentiation. dedicated stakeholder management is required.
• Investments in AI partner platforms: TechMahindra brings the IBM Watson portfolio to clients by • Demonstrate industry relevance, not just technical
Plan using its AQT automation framework and toolkit, reiterating the emphasis on industrialized experience: Clients encourage TechMahindra to help
services. It has also invested in building joint use-cases on Leonardo, with SAP co-innovation them envision the future of their business. While they
Implement labs. appreciated technical capabilities, they would like to
• Broad set of marquee clients: Despite accelerating its journey in automation and AI, see the demonstration of industry experience next to
Manage data scientists.
TechMahindra can reference marquee clients demonstrating the viability of its automation
strategy. • Expanding toward project-centric work: In order to
Operate • Willingness to convey and share best practices across its client base: Clients called out that reinforce client relationships and progress toward
TechMahindra facilitated discussions among its clients around best practices and lessons learned higher value engagements, TechMahindra should
Optimize opportunistically invest in more project-centric work.
from the early deployments.
Recent acquisitions include: Tech Mahindra works with clients across Geographic footprint and scale of the AI practice: • AQT (Automation, Quality, Time) : Automation
• FixStream Network (strategic industry sectors: Platform consolidates their automation solutions,
stake 2014) The IA Practice comprises of a team of 500 practices and tools.
• Leading American associates; 350 of them are centrally invested in the • Tactix: AI (NLP, ML, deep learning) and analytics-
Tech Mahindra has a broad set of telecommunications provider Center for Automation Technologies and another powered IT helpdesk and operations automation
partnerships, including: • An oil and gas subsidiary of a large 150 are from various competences and start-ups that platform.
• IBM Watson American conglomerate Tech Mahindra is funding. • Meridian: Visualization and correlation for single-
• Microsoft • Large UK-based telecommunications Central teams own the automation strategy and pane-of-glass view of apps and infrastructure
• IPsoft provider enablement across all service-lines and are involved through patented technology Flow2Path and
• Arago • Canadian multinational bank in the IP development and implementation visualizations.
• Pega • • Entellio: Enterprise chatbot platform for enhanced
North America based manufacturer in enablement for AQT platforms. They manage the
• SAP customer experience.
the aviation space Alliances and Partnership centrally and house three
• Epselon: Digital enablement and analytics
• Cortex • Large European manufacturer in the centers of excellence for service automation, RPA,
platform for SAP and Oracle EBS.
• Avaamo aviation space and AI. • LitmusT: AI and analytics-powered test
• Global pharmaceutical company AI CoE is focused on developing use-cases across key automation platform.
• Canadian multinational financial industry verticals like Telco, BFSI, • Acumos: AI platform for building and managing
services company Manufacturing/Engineering and Healthcare. About ML/DL models and market place; joint investment
• Global healthcare provider 250 engineers who are certified in AI streams spread with a large telecommunications client.
• Leading theme park in APAC across their major delivery centres across the globe.

29j916d5
Wipro Global service provider driving industrialization by expanding its
flagship HOLMES investment toward broader partner ecosystem
High Performer
• Scale and repeatability of • Focus on the industrialization of automation and AI: Wipro’s early and expansive investments in the • Investments in marketing and go-to-market:
deployments HOLMES platform and framework provide a focal point for marketing and go-to-market initiatives. HOLMES is being undersold. HOLMES was one of the
• Data management strategies first broad automation frameworks on the market.
• Acceleration of eco-system mindset: Wipro is evolving from its early emphasis on proprietary IP to
supporting integration of The narratives don’t reflect the deep capabilities,
embracing a much broader and open ecosystem of partners. Reference points are partnerships with
semi and unstructured data including neural networks and NLP, but Wipro is
• Vision for and investments Arago, Avaamo but also some of the platforms of its competitors. The intent is to be able to offer more
modular services all the way to Automation-as-a-Service. The same applies for leveraging the leading starting to address this. More importantly, the
in the evolution of
RPA tool sets as an integration into HOLMES, thus aligning Wipro with the mainstream RPA scenarios. expansive set of assets in the Bot Library are not well
Intelligent Automation
• Actual delivery of services • Aligning automation and data platforms: Wipro’s Data Discovery Platform (DDP) is providing the communicated. Too much emphasis has been placed
foundational data layer for HOLMES. Wipro’s emerging ‘Data as IP’ practice is focused on building on use cases such as KYC and Autonomics.
vertical and horizontal ontologies and data sets that can be leveraged for use cases directly. • Drive automation and AI through the organization:
Plan • Careful buildout of vertical use cases: Wipro is carefully building capabilities around specific use cases Aligned with the lack of marketing, some clients
in order to demonstrate robust delivery. An initial strong focus is on KYC and the broader BSFI vertical. encourage Wipro to educate its executives on the
Implement Other use cases include Case Law Management, FACTA Tax Forms, and Contract Processing. However, capabilities of HOLMES as they don’t always come
the Bot Library contains an impressive array of assets across the entire IA Continuum. Examples are the across proactively and effectively.
Manage Natural Interaction Bot, Cognitive Insights, and Code Generation Bot. Other innovations include a
Classifier Bot deployed on AWS that can be integrated into ServiceNow. • Expansion of consulting capabilities: Some clients
Operate encourage Wipro to demonstrate industry
• Leverage of Design Thinking: Leveraging acquisitions such DesignIT and Cooper. Wipro advocates joint
relevance, not just technical experience. While
discovery and development with customers in its new BOT Studio. This approach also focuses on Wipro is starting to build an Automation Advisory
Optimize external collaboration with start-ups. unit, it should drive AI into broader engagements.
Partnerships include: Wipro works with clients across industry sectors: Geographic footprint and scale of the Artificial • Frameworks within HOLMES include: Automation
Intelligence practice: Studio: One stop to view, create, publish and
• Google • US multinational banking and financial services consume CPA and RPA solutions; Deep Extraction
• Microsoft company Having been incubated in the CTO Office for Framework for Text: linguistics and deep learning
• AWS • Leading US investment banking and financial the initial years, Wipro HOLMES and based text extraction framework; Image Processing:
• IBM Watson services corporation Automation Portfolio is now a dedicated unstructured document text/image processing
• Avaamo • British multinational construction company practice to drive Wipro's AI and Automation through OCR, NLP, and Machine Learning; Chat:
• Arago • Leading Europe-based oil & gas company charter. Supported by a team of 2,000+ smart virtual assistant to improve the productivity
• Tricentis • European multinational focused in the areas of resources dedicated to HOLMES and and efficiency of workforce; Cognitive Search:
electronics, healthcare and lighting Automation Ecosystem as well as across information extraction and understanding
• American multinational information technology technology and business units, Wipro’s AI framework to support smart queries; Mimictron:
Mimic user behavior through goal oriented actions
company practice has a global footprint with geo teams
using Deep Learning.
• Leading European IT and consumer electronics in US, Europe and APAC. Wipro has trained • Bot Library (Curated Bots): Standardized and
company over 4,000 resources on generic AI/ML skills to productionized bots ready for rapid deployment to
• U.K based multinational oil and gas company cross-leverage across engagements. cater to pre-identified use cases. These bots from
• German multinational automotive corporation the Wipro HOLMES Bot Library require minimal
• Financial services company in Australia and customization, e.g., eKYC (validating customer
New Zealand information in banks), Classifier Bot (NLP-based
ticket categorization, allocation, and resolution).

29j916d5
LTI Global service provider focusing on platform approach and accelerating
High Potential the buildout of proprietary knowledge and AI assets
• Vision for and investments • Renewed corporate strategic focus: The IPO of L&T Infotech in 2016 has ended an era of change • Invest in brand and stakeholder management: As
in the evolution of and lack of strategic clarity. LTI has refined its brand and strategic direction under the initiative LTI LTI is accelerating its efforts around its LTI 2.0
Intelligent Automation 2.0. This allows for more certainty in decision making in particular when it comes to innovation initiative, it needs to double down on conveying its
• Leverage of Design Thinking topics such as AI. Clients reference that they like LTI’s focus on business value. Similarly, they strategic intent to the broader market. The turbulent
• Governance and testing highlighted that LTI is happy to go beyond scope as they are looking to position themselves for a last couple of years require a clearer articulation of
services larger share of wallet. its strategic bets, in particular as the market moves
• Training and supporting toward the secular shift of digital and automation.
clients • Focus on platform approach: LTI is focusing its investments on a platform approach. Its Mosaic AI
platform is built with an ecosystem mindset using open source and mega ISVs as its key building • Consistency and leverage of data assets: LTI should
blocks. The platform is offering in excess of 50 microservices including speech to text, vision explore finding efficiency across its data assets. The
services, and ticket classification, which is underpinned by an ever increasing taxonomy and cross fertilization with analytics efforts could be
Plan knowledge base that forms its IP. The approach is rooted on a core belief that AI has to be about enhanced. LTI is conscious of not wanting to be
enhancing human lives. LTI believes this strategy will work as it is seen with an end-to-end dependent on data requirements, hence the focus is
Implement perspective of solving problems by plugging in intelligence into systems and context. The Mosaic AI to build out domain-specific assets.
platform eco-system approach seems to work. To accelerate this, it is building out an academy as • Align AI with broader Intelligent Automation
Manage well as online training. strategy: LTI has been leveraging Intelligent
• Formalized AI practice: Albeit nascent and small in size, LTI has as a formalized its Artificial Automation as a differentiator to win large deals.
Operate Intelligence and cognitive practice. This focus has led to a solid strategic vision. However, LTI has to Continuing this trend they should align AI stronger
demonstrate that it can scale this vision and drive it through the organization. Thus far, has focused with the broader automation requirements.
Optimize on industrialized services and been more conservative around project-centric requirements.
Acquisitions include: LTI works with clients across industry Geographic footprint and scale of the AI • MOSAIC AI platform: Helps companies accelerate operational
• AugmentIQ sectors: practice: digitization by building cognitive capabilities into their existing
knowledge assets. It deploys self-learning systems to deliver
Partnerships include: • Large midstream oil and gas Artificial intelligence and cognitive
intelligent and contextual content to all users. The platform
• Google company practice has been focusing on multi-
• Microsoft • Niche medical device company disciplinary skills and has hired talent from includes responsive interfaces, intelligent “bots”, and NLP
• AWS • Large media and entertainment some of the leading educational institutes capabilities to transform legacy systems.
• IBM Watson major in India like the IIT and Indian Statistical Solutions IP built on the platform include:
• Workfusion • Large insurance broker Institute. This includes skills in • Cognitive Claim: Faster claims processing; integrating deep
• Large UK-based bank in Singapore programming, statistics, mathematics, and learning services like IBM Watson Conversations and Vision.
Technologies include: • Multinational CPG manufacturer algorithm development around NLP and • NextGen Service Desk: A self-service chat bot with intelligent
• Stanford NLP • Large UK based bank in South vision. They focus on a design thinking
ticket classification, smart routing, and auto resolution.
• NLTK Africa approach to solve problems and have
• OpenNLP • Leading US global bank dedicated labs and experience centres to • Financial Crime EDD Automation: Intelligently automates
• Scikit • Leading Japanese automotive enable innovation and IPs being creation of EDD document from unstructured data with entity
• DarkFlow manufacturer developed. Consists currently of around co-referencing and scoring against financial crime exposure
• TensorFlow • Leading network device 1560 FTEs and focuses on IP driven categories.
• OpenCV manufacturer approach to solve narrow problems and • Trade Finance Automation: Automation of extraction of
• Keras address globally. entities and compliance screening for assisted operations AI
and automation.

29j916d5
Market Predictions and
Recommendations
29j916d5
Market Predictions in Enterprise AI Services
 Organizational maturity around AI to accelerate: Most providers will either understand AI as a
foundational layer across their organization or even move to a dedicated AI practice. This increasing
maturity will lead to a more robust education of stakeholders.
 Stack players and mega ISVs will double down on AI: Oracle’s announcement of its Autonomous Database
as well as SAP buying Recast.ai are an indication for an arms race on AI by the large stack players and mega
ISVs. Those capabilities will be blended into both existing offerings and enhance specific AI platforms such
as SAP Leonardo.
 Most startups will be absorbed by M&A: As one supply side executive pointedly put it, “the
understanding of what is best-of-breed is changing weekly”. Therefore, providers, especially outside the
Winners Circle, are coy to devote too many resources to partner with those startups. However, a set of
providers like Avaamo, CognitiveScale, and Loop AI have the opportunity to evolve into a reference
partner for broader AI capabilities.
 New set of specialist consultancies coming to the fore: Just like Genfour and Symphony Ventures were at
the forefront of educating and leading transformational RPA projects, we are likely to see the emergence
of a new breed of consultancy replicating. Some of the RPA pure plays will morph into broader offerings.
 Lack of governance might lead to high profile litigation: The absence of discussions around risk and
immature governance strategies that underestimate the impact of self-learning and self-remediating
systems is likely to lead to high-profile cases of litigation, which in turn will help to mature AI capabilities.
 Automated training of Machine Learning algorithms will accelerate and transform the market:
Technologies such as Gluon and AutoML are accelerating the training of algorithms. As slide 25 has
highlighted, the Holy Grail of AI is at the intersection of iterative data inputs and minimal training of
algorithms. Higher levels of unsupervised learning are critical to advance toward notions of end-to-end
processes to accelerate the journey toward the OneOffice.
29j916d5
How Service Providers Can Differentiate in 2018 & Beyond
 Demonstrating the impact on the Future of Work: As the data on slide 62 shows, organizations see the
biggest potential for AI in either augmenting or replacing labor. Therefore, a realistic guidance as to how
organizations have to approach the transformation of their workforces is essential. This should include HR
strategies, talent management, and effective change management.
 Clarity on required investments whilst supporting change management: As the data on slide 63 outlining the
inhibitors for AI suggests, the most critical issues preventing AI projects are demonstrating that the required
investment level justifies any potential disruption and redresses the notion that enterprise software
platforms will provide all the necessary capabilities over the next five years.
 Guidance on use cases and lessons learned: The fourth most critical inhibitor is that organizations don’t
know where to start. Just like the early RPA days, the market urgently needs realistic guidance and thought-
leadership. Of critical importance is playing back the experiences from the early deployments.
 More nuanced communication between industrialization and project-centric requirements: As we have
called out repeatedly, the starting points and context for AI projects can be vastly different. Thus, go-to-
market strategies have to expanded by more nuanced communication taking into account the respective
vantage point of stakeholders.
 Governance beyond data: As organizations are progressing toward the OneOffice with ever more scalable
end-to-end processes, the market urgently needs more holistic approaches to governances that brings
security, testing services, process, and data governance together. GDPR legislation is a telling example for the
urgency in building out those capabilities.
 Put the process owner center stage: Avoid jargon and technology-centric communication and put the process
owner center stage by discussing business problems being solved in ways that are not possible with other
technology and approaches.

29j916d5
The Biggest Benefits of AI
What do you believe to be the greatest potential benefit of AI?
Source: “State of Automation 2017”

Sample: n=181 Buyers of AI

29j916d5
What Is Preventing the Use of AI?
What is stopping you from using AI Automation?
We don’t believe the cost savings would offset the

32%
disruption
It will be built into enterprise software platforms in the
30%
next 5 years we’ll wait
Our underlying platform is good already and we don’t
29%
think Automation will deliver any lower operating costs
We are not sure where to start 28%
We don’t see the business value to our organization 23%
We cannot get the funding 21%

We have adopted technology-driven process change
21%
before and have had bad experiences
It’s just hype 18%
We are unconvinced of the technology in this area 16%
We don’t know who should be responsible 11%

We are unconvinced of the service provider capabilities
10%
in this area
Providers don't offer enough risk / gain-sharing to
8%
support outcomes
The disruption to the business is too great 7% Sample: n=219 Non-Buyers of AI

29j916d5
2018 Recommendations: Service Providers
 Invest in guiding clients around business problems and use cases: Move beyond tool and technology
mindset in order to support clients’ journey toward the OneOffice. In terms of communication, best practices
are approaches by firms like Capgemini and KPMG to describe the services delivered around generic activities
including extract, interpret, classify, diagnose, retrieve, and recommend. Furthermore, there is an acute lack
of realistic thought leadership on Enterprise AI, thus by being seen as an innovator and thought leader you
can achieve competitive differentiation. By doing so, don’t fall into the trap of highlighting unicorns and
consumer technology issues. The relevant context should be enterprises held back by legacy systems.
 Drive AI as a foundational layer across the organization: To advance toward the OneOffice, progress on
collapsing organizational boarders and drive AI and data as a foundational layer across the organization with
the end goal of straight-through processing and end-to-end processes. Help clients to understand that they
need the blend of the skills sets of data engineers and data scientists in order to be able to directly integrate
data into production.
 Explore the industrialization of data science: As the market is awash with lower level machine learning and
conversational services, the industrialization of high level and high value data science might provide both
differentiation and stickiness in client relationships. In other words, leverage and blend both industrialization
and project-centric work.
 Re-evaluate partner eco-systems: IBM Watson, TCS ignio, and Infosys NIA are examples for ecosystems that
are being opened to competitors. Explore co-innovation and be clear about your strategic bets. In particular,
demonstrate domain-specific capabilities and use cases. This can help to overcome fears of vendor lock-in
around data issues that are widespread among the buy side.

29j916d5
2018 Recommendations: Buyers
 Prioritize the evaluation of processes that can be automated through AI: Building on the learnings from
the early RPA deployments, prioritize the evaluation of processes that can be automated and enhanced by
AI. Be clear on the business goals that are meant to be achieved.
 Make talent a strategic priority: Talent that understands the innovation subsumed under the AI moniker
and more importantly their impact on workflows and process chains is scarce. Critically the buy side won’t
have first (or even second) pick from graduates. Therefore, organizations should review and refine their
strategic approach to talent.
 Leverage Design Thinking and co-innovation: The ultimate goal of your AI projects should be to solve
problems that can’t be solved with other technology and approaches. Focus on re-imagining your
processes rather than applying AI as patch work. Seek out specialist consultancies that can help drive the
process.
 Make data the cornerstone of your delivery strategy: The Holy Grail of service delivery is at the
intersection of automation, analytics, and data, particularly around the blending of iterative data inputs
with minimal training of algorithms. Be mindful of the hype and hollow promises that platforms, cognitive
libraries and machine learning will deal with your business problems. Rather invest in talent that blends
data engineering and data science.
 Move beyond an FTE-centric mindset: Mature buyers are progressing from an FTE-centric mindset and
traditional approaches to business cases. The essence of digital and Intelligent Automation is that routine
service delivery is being decoupled from labor arbitrage. This necessitates to move to softer metrics such
as business agility and customer satisfaction. The expectation is that on high scores for those metrics, cost
savings and other tangible benefits will follow.

29j916d5
About the Author
29j916d5
Tom Reuner
• SVP, Intelligent Automation and IT Services, HfS Research – London, England
Overview
▪ Tom Reuner is Senior Vice President, Intelligent Automation and IT Services at HfS. Tom
is responsible for driving the HfS research agenda for Intelligent Automation and IT
Services. Automation cuts across the whole gamut ranging from RPA to Autonomics to
Cognitive Computing and Artificial Intelligence. This includes increasingly the
intersections of unstructured data, analytics, and Cognitive Automation while
mobilizing the HfS analysts to research Intelligent Automation dynamics across specific
industries and business functions. Furthermore, he is supporting HfS’ push to disrupt IT
Services research by focusing on application services and testing. A central theme for all
his research is the increasing linkages between technological evolution and evolution in
tom.reuner@hfsresearch.com
the delivery of business processes.
Previous Experience
▪ Tom’s deep understanding of the dynamics of this market comes from having held
senior positions with Gartner, Ovum and KPMG Consulting in the UK and with IDC in
Germany where his responsibilities ranged from research and consulting to business
development. He has always been involved in advising clients on the formulation of
strategies, guiding them through methodologies and analytical data and working with
clients to develop impactful and actionable insights. Tom is frequently quoted in the
leading business and national press, has appeared on TV, and is a regular presenter at
conferences.
Education
▪ Tom has a PhD in History from the University of Göttingen in Germany.

29j916d5
2018 AI Research Agenda

29j916d5
The HfS Mission: Defining Future Business Operations
HfS’ mission is to provide visionary insight into the major innovations impacting business
operations: automation, artificial intelligence, blockchain, digital business models, and smart
analytics. We focus on the future of operations across key industries. We influence the strategies
of enterprise customers to develop operational backbones to stay competitive and partner with
capable services providers, technology suppliers, and third-party advisors.
HfS is the changing face of the analyst industry combining knowledge with impact:
• ThinkTank model to collaborate with enterprise customers and other industry stakeholders
• 3000 enterprise customer interviews annually across the Global 2000
• A highly experienced analyst team
• Unrivaled industry summits
• Comprehensive data products on the future of operations and IT services across industries
• A growing readership of more than one million annually.
The As-a-Service Economy and OneOffice™ are revolutionizing the industry.

29j916d5

RS 1803 HFS BP Enterprise Artificial Intelligence (AI) Services 2018

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

RS 1803 HFS BP Enterprise Artificial Intelligence (AI) Services 2018

Uploaded by

Copyright:

Available Formats

The Services Research Company

HfS Blueprint Report

Enterprise Artificial Intelligence (AI) Services 2018

Introduction and Key Definitions 2

The State of the Enterprise Artificial Intelligence Market 10

Service Provider Grid 32

Service Provider Profiles 38

Market Predictions and Recommendations 59

About the Author 66

© 2018 HfS Research Ltd. Proprietary │Page 1

 While the broader market appears stuck in

© 2018 HfS Research Ltd. Proprietary │Page 3

Technology: Nuance, Cortana, Alexa

Neural Technology: HIRO, ignio, HOLMES

Technology: IBM Watson, Amelia, LivingActor, Accenture myWizard

Technology: IPcenter, HIRO, ignio, HOLMES

Computer Technology: AntWorks, AlchemyAPI, Clarifai

© 2018 HfS Research Ltd. Proprietary │Page 5

Intelligent Automation Value Chain

Plan Implement Manage Operate Optimize

© 2018 HfS Research Ltd. Proprietary │Page 6

© 2018 HfS Research Ltd. Proprietary │Page 8

© 2018 HfS Research Ltd. Proprietary │Page 11

While nothing in the context of automation

© 2018 HfS Research Ltd. Proprietary │Page 12

Service orchestration Specific requirements

Horizontal Sub-sector lens

Out-of-the-box Design Thinking

© 2018 HfS Research Ltd. Proprietary │Page 15

© 2018 HfS Research Ltd. Proprietary │Page 16

© 2018 HfS Research Ltd. Proprietary │Page 17

Exploring quantum computing techniques on AI and Machine Learning

Leverage of ML for medical coding: Global SI (not Watson) helped European

Cross-fertilization from other sectors: Global SI has helped Australian company to

© 2018 HfS Research Ltd. Proprietary │Page 18

© 2018 HfS Research Ltd. Proprietary │Page 19

© 2018 HfS Research Ltd. Proprietary │Page 20

© 2018 HfS Research Ltd. Proprietary │Page 21

© 2018 HfS Research Ltd. Proprietary │Page 22

Source: HfS Research “State of Automation 2017”

© 2018 HfS Research Ltd. Proprietary │Page 23

 The Holy Grail of AI is at the

© 2018 HfS Research Ltd. Proprietary │Page 24

 The difference between Big Data projects and

 Machine learning requires “learnable” data.

© 2018 HfS Research Ltd. Proprietary │Page 28

© 2018 HfS Research Ltd. Proprietary │Page 29

© 2018 HfS Research Ltd. Proprietary │Page 30

© 2018 HfS Research Ltd. Proprietary │Page 31

© 2018 HfS Research Ltd. Proprietary │Page 33

© 2018 HfS Research Ltd. Proprietary │Page 34

© 2018 HfS Research Ltd. Proprietary │Page 35

Expansion data science

Strong = Most Maturity

Modest = Least Maturity

© 2018 HfS Research Ltd. Proprietary │Page 36

Expansion data science

Strong = Most Maturity

Modest = Least Maturity

© 2018 HfS Research Ltd. Proprietary │Page 37

Full Value Chain offered “Build” not offered

© 2018 HfS Research Ltd. Proprietary │Page 39

© 2018 HfS Research Ltd. Proprietary │Page 40