Enterprise AI Guide

The Executive’s Playbook

for Business Success with AI
INTRODUCTION TO AI AND ML

Artificial intelligence is quickly becoming a

fundamental building block of business operations,
providing real gains such as improved processes, 40%
of digital transformation initiatives
increased efficiency, and accelerated innovation. will be supported by AI in 2019
Advances in machine learning technology have
combined with high-performance compute options
and an abundance of data to create a perfect storm
for AI to transform organizations of all sizes.
42%
of executives say AI will drive
innovation in their organization
For enterprise leaders, the need to embrace AI capabilities is immediate.
IDC estimates that 40% of digital transformation initiatives will be
supported by AI in 2019¹ and 42% of executives say AI will drive innovation
in their organization.² As enterprises accrue data from numerous sources—
ranging from customer feedback, sales logs, internal processes, and financial
forecasting—there is opportunity to operationalize this often “dark data” into
insights that can improve decision-making, create new revenue streams, and
optimize the business value chain.

¹ “Worldwide Spending on Cognitive and Artificial Intelligence Systems Will Grow to $19.1 Billion in 2018,” March 22, 2018, IDC
² “Reworking the Revolution: Future Workforce,“ 2018, Accenture

2
2
GOALS

Impacting business goals

Early adopters are weaving AI across the organization
to address business priorities. Some machine learning
projects tackle incremental gains to automate processes
to create efficiencies. Others are transformational
initiatives aimed at innovation and competitive
differentiation. While there are numerous applications
of AI, enterprises already realize great value from use
cases that provide new experiences for their customers
and drive business growth.

Goals for AI innovation

1 Enable product and service innovation

2 Drive research and discovery

3 Enhance the customer experience

4 Improve customer service

5 Improve healthcare outcomes

6 Increase efficiency and productivity

7 Improve security and compliance

8 Optimize supply chain operations

9 Improve decision making

3
GOALS

1 Enable product and service innovation 3 Enhance the customer experience

While finding new ways to innovate is top of Across industries including financial services,
mind, more mature AI enterprises already use healthcare, retail, and media and entertainment,
machine learning pervasively. They inform their targeted recommendations have become important
product roadmap through intelligence extracted for personalization that delights and engages
from customer-feedback; drive the product customers. Epsilon research indicates 80 percent
development lifecycle, including DevOps and of consumers are more likely to make a purchase
quality assurance through automation and when brands offer personalized experiences.
intelligence; and infuse machine learning Using voice-of-customer analytics to understand
capabilities directly into new products that customer behavior and preferences, as well as
benefit the end user. recommendations engines to predict best matches,
enterprises can push customized offers to
location-based mobile services that offer real-time
2 Drive research and discovery convenience and choice.

AI-driven research unlocks new ideas in diverse

industries such as automotive, manufacturing,
and energy and utilities, where machine learning
capabilities help uncover new formulas, new
materials, and new ways of thinking about age-
old problems. It is especially relevant in industries
such as healthcare and life sciences, where
machine learning models can make predictions
that drive research toward breakthroughs that
lead to new and improved patient treatments.

4
GOALS

4 Improve customer service 5 Improve healthcare outcomes 6 Increase efficiency and productivity
Critical to every business is ensuring that Today AI is being used in healthcare to help Across business operations, pervasive use of
various customer touchpoints—both internal patients quickly locate available in-network machine learning to automate and streamline
and external—provide timely, accurate, and doctors. Behind the scenes, machine learning is processes has resulted in increased efficiency and
meaningful customer service. AI technology being used to analyze everything from X-rays reduced costs. For example, using automated media
such as conversational agents bridge the gap, to patient-history data. As well, it can make tagging and large-scale document recognition and
modernizing the contact center to provide quick predictions on patients at risk, and help physicians analysis, machine learning can eliminate countless
responses to customer calls and chat requests. and hospital personnel determine the best hours of labor. Machine learning can also analyze
In addition, text-to-speech and natural language course of action. manufacturing incidents in real time, providing early
processing capabilities can be applied to the warning of potential problems. HR departments
streams of incoming customer data such as are even turning to machine learning to assess job
texts, voice messages, and customer service applicants more efficiently.
logs to better understand customer needs and
sentiment in order to help improve the quality
of customer service.

5
9 GOALS

7 Improve security and compliance 9 Improve decision-making

Use of machine learning across security, risk, and Machine learning-based predictions augment
compliance use cases is a fast-growing trend decision making across all functions of the
especially in the financial services sector. organization and across use cases. These predictions
Fraud-detection models help to keep consumer produce continuous actionable insights that help
data safe and prevent malicious attacks against leaders and teams tackle operational and business
bank accounts and even mobile endpoints. challenges. This can range from forecasting that
On the other end of the spectrum, machine impacts pricing optimization, to predictions for
learning automates mundane tasks such as autonomous vehicles to make informed choices.
financial-document analysis, reducing manual
effort and allowing the workforce to focus on
higher-value tasks.

8 Optimize supply chain operations

From point-of-sale to freight-delivery management
to forecasting demand, AI plays a pivotal role
across the supply chain. In demand planning
for example, machine learning algorithms can
look at historical data as well as demand and
other relevant data—such as product schedules,
weather patterns, and competitor pricing—to
determine when it’s time to restock or end of life
a product. Sophisticated machine learning models
can predict weekly, daily, and even down-to-the-
minute forecasts to help companies streamline
supply chain management, save cost, and
increase efficiency.

6
CHALLENGES

Challenges of adoption
Machine learning is frequently the catalyst that turns business data into prepping data, choosing algorithms, building, training, and deploying models
accurate predictions and actionable information, but as with many emerging … and iterating over and over again. There are decisions to be made about
technologies, there are challenges to adoption including data, complexity, infrastructure—selecting the right compute for training and inference,
cost, and lack of skills. considerations for cloud, on premise, and edge deployments.

Data ambiguity: Enterprises can struggle with various issues related to Cost inhibitors: Machine learning training and inference can be expensive,
data. First and foremost, many are unaware of all their possible data sources especially since models require iterations to improve the accuracy of predictions.
that may hold hidden insights. Even when they’ve identified data, there’s a Because embarking on AI initiatives is new to many companies, they also
lack of labeled data ready for machine learning. Furthermore, even labeled don’t have the experiences or skills in-house, and often have to rely on external
data can prove to be an issue where integrity is in question since data can resources to kick-start projects.
often have hidden biases from human labelers. Finally, enterprises often
struggle with ensuring the right data management and governance policies Lack of skills: Even when companies embrace new technologies like AI
are in place to allow the right people and processes to securely access, store, to drive business transformation, not having the right skills is often a road
and manage the data. block to getting started. AI projects require machine learning expertise to
build and train ML models—this includes the skills of machine learning
Complexity: The machine learning workflow can be time-consuming and developers, data scientists, and researchers to build algorithms and train
iterative, which leaves many organizations and developers thinking machine models. These skills are not in great supply and are often unavailable in-house
learning is complex and difficult to use. There are many steps involved from which leaves businesses struggling to train or supplement their skills with
consultants and partners.

7
KEY CONSIDERATIONS

Getting started with AI Key considerations

Understand business objectives

AI presents new opportunities to realize foundational

Advance your data strategy
gains such as efficiency and cost savings, as well
as higher-value gains such as product innovation Leverage the cloud

and spurring discovery and research. But how do Enable your organization
organizations get started? For many, AI adoption
begins by identifying workflows and business
processes that suffer from low efficiency or where
human mistakes abound. They consider all their data
sources and existing data strategy. They determine
the best cloud-based infrastructure and tools to scale
AI. And last, they ensure that the right skills are on
board for machine learning projects to be successful.

8
KEY CONSIDERATIONS

Identify desired outcomes and

deliver better experiences for
your customers

Understand business objectives

Understanding the business benefits of AI adoption—in particular, the specific • Understand the impact: From the outset, consider the operational effect
benefits relevant to your organization—is critical for enterprise success with AI. of new AI solutions. AI can have transformative impacts, so it’s important
Once objectives are identified, it’s important for business and technical leaders to plan in advance for what you want to achieve and measure. This can
to understand and champion their role. also help you determine how to measure success. When considering ROI,
ensure you have value checkpoints early in the project lifecycle. This lets
• Select a targeted use case: When choosing a pilot, consider use cases you adapt the approach before you’ve scaled investment.
where AI can have the most impact, and those from which you can learn
to scale enterprise deployment. Focus on how you can deliver a better • Iterate and learn: Once you’ve proven the potential of machine learning,
experience for your customers and identify the business and operational the next step is to move from pilot to production, which may include
outcomes desired. Then establish one or two high-value proofs of concepts integrating the machine learning capability into a larger IT system.
(PoC) that can really make a difference to your organization and quickly This move typically takes longer than the pilot process and can vary
demonstrate results. For the PoC to succeed, it’s critical to have the right depending on the complexity of the overall system and how large-scale
resources in place, including infrastructure, data, and capabilities. the production deployment will be.

9
KEY CONSIDERATIONS

Advance your data strategy

Data is gold for leaders who are looking to disrupt their industries with
machine learning, but many organizations don’t have machine learning-ready
data. Recognizing the importance of data and developing a plan to collect
and use that data is critical for successful machine learning adoption, even at
the PoC stage.

All sources of data need to be uncovered, from structured data like billing
and CRM to unstructured data like social media feeds, images, and forms.
Then that data needs to be evaluated for quality and usefulness. Finally, data
needs to be cleaned and accurately labeled for machine learning models to
transform it into valuable insights.

Leverage the cloud

Successful machine learning initiatives need more than just the right tools.
A comprehensive platform brings together data store, security, and analytics
services, as well as compute resources for training and deployment. Turning to
the cloud for these services brings a wide range of benefits, including speed,
scalability, flexibility, resilience, security, and reduced cost.

As well, the cloud offers the widest range of high-performance CPU and
GPU processor types, which are essential for large-scale training and for
deployment in a production environment. Using cloud-based data lakes and
storage also ensures that you can easily access and manage data so that
machine learning initiatives are seamless, repeatable, and scalable. Data is gold for leaders who are
looking to disrupt their industries

10
KEY CONSIDERATIONS

Enlist the right stakeholders

Enable your organization and skills to get pilots off
the ground and scale
Along with the right use cases, having the right skills to build machine learning
applications and systems, as well as the right process and operating model,
are essential to getting pilots off the ground and scaling enterprise AI.

• Assemble the team: Consider appointing a Chief Data Officer (CDO) to lead
the charge on data strategy and governance, bring together interdisciplinary
teams, and streamline data processes. Assemble a team of machine learning
developers and data scientists essential for a successful PoC, and train teams
for future deployments. It’s also important to involve subject matter experts
who understand your business vernacular, especially for industry domains,
to help you get to ground truth with your data.

• Create the process: AI may not bring the expected value if the results
are not integrated with other areas of the organization. Operationalizing
machine learning models is hard—as many as half of PoCs don't get
deployed into production. Therefore executive sponsorship to change
business processes and alignment with applications development is key.
Successful teams create processes to align AI experts, data scientists and
developers with key business stakeholders. A well-defined process also
helps ensure the final output is well integrated into business processes.

• Build the culture: To help realize its potential, there needs to be cultural
acceptance that AI is an important part of business and operations.
Some initiatives may require information from across these domains, so
it’s important to understand all the stakeholders who need to be involved,
and bring together stakeholders who can champion adoption.

11
GET STARTED WITH AWS AI

Getting started with AWS AI

AWS has the broadest and deepest set of AI and Machine Learning services
for your business. AWS and Intel have been working together for over a
decade to deliver the most comprehensive set of resources, tools, training,
and services. On behalf of our customers, we are focused on solving some
of the toughest challenges that hold back AI adoption.

Choose from pre-trained AI services; Amazon SageMaker to build and scale

machine learning; or build custom models with support for all the popular
open-source frameworks. AWS AI capabilities are built on a comprehensive
cloud platform, optimized for machine learning with high performance
compute, security and analytics.

AI Services: AWS pre-trained AI Services provide ready-made intelligence

200+ 75%
for your applications and workflows. AI Services easily integrate with
your applications to address common use cases such as personalized
recommendations, modernizing your contact center, improving safety
and security, and increasing customer engagement. Because we use New ML features and Lower ML inference cost
services since Jan 2018
the same deep learning technology that powers Amazon.com and our
Machine Learning Services, you get quality and accuracy from continuously-
learning APIs. And best of all, AI Services on AWS don't require machine
learning experience.

ML Services: Amazon SageMaker provides every developer and data

scientist with the ability to build, train, and deploy machine learning models
quickly without needing data engineers or DevOps. Amazon SageMaker,
One-click
Model training
10X
Better ML algorithm
with optimizations from Intel, is a fully-managed service that covers the and deployment performance
entire machine learning workflow to label and prepare your data, choose
an algorithm, train tune and optimize models for deployment, make
predictions, and take action. Your models get to production faster with
much less effort and lower cost.

12
GET STARTED WITH AWS AI

Amazon ML Solutions Lab: The Amazon ML Solutions Lab combines hands-

on educational workshops with advisory professional services to help you
"work backwards" from business challenges, and then go step-by-step through
the process of developing machine learning-based solutions. You'll be able to
take what you have learned through the process and use it elsewhere in your
organization to apply machine learning to business opportunities.

AWS Machine Learning Competency Partners: AWS Machine Learning

Competency Partners offer a range of services and solutions to help you create
intelligent solutions for your business. From concept development through
execution, partner solutions offer tools and expertise to help with the entire
machine learning workflow.

Training and Education: Start training on machine learning on AWS with

courses based on the same material used to train Amazon’s developers
through the combination of foundational knowledge and real-world
application. Developers, data scientists, data platform engineers, and business
decision makers can use this training to learn how to apply machine learning,
AI, and deep learning to their businesses.

To learn how to put AI to work for your

business, visit AWS.ai

13
PARTNER SHOWCASE

Build your organization’s AI with AWS and Intel

More machine learning is built on AWS than anywhere else, and getting started has never Amazon SageMaker enables one-click training,
been easier. model tuning, and one-click deployment to the
cloud or edge aws.amazon.com/sagemaker
AWS provides a smooth and simple on-ramp for companies looking to adopt AI and ML:

Amazon EC2 C5 and C5n instances offer the

• Support for all major ML frameworks, including Intel-optimized versions of Tenser
Flow and MXNet, giving engineers and developers built-in performance and freedom lowest price per vCPU in the Amazon EC2 family
of choice and are ideal for running advanced compute
intensive workloads like ML training and inference
• Pre-built algorithms and model creation tools with Amazon SageMaker, including aws.amazon.com/ec2/instance-types/c5
reinforcement learning tools like Intel® RL Coach and Ray RLib
AWS DeepLens is a fully programmable video
• Partner-built models in AWS ML Marketplace that can quickly get your organization camera, tutorials, code and pre-trained models
started with the most common use cases designed to expand deep learning skills
aws.amazon.com/deeplens
• Optimized infrastructure–such as the Amazon EC2 C5 and C5n instances –running on
custom Intel® Xeon® Scalable Processors for top performance on compute intensive ML AWS DeepRacer gives developers a way to
training and inference workloads
get hands-on with machine learning, experiment,
and learn through autonomous driving
• Software and hardware tools – including AWS DeepLens cameras and AWS DeepRacer –
aws.amazon.com/deepracer
featuring Intel Atom® processors and Intel OpenVINO™ software that empower
your developers to learn about ML hands-on
AWS ML Marketplace – over 150 pre-built
models and algorithms– including 14 from
• AWS and Intel offer a full portfolio of business development and technical training
resources, giving your organization the skills it needs to push AI forward with confidence Intel–to get you started on the most common
ML projects aws.amazon.com/marketplace/
solutions/machinelearning

14
© 2019, Amazon Web Services, Inc. or its affiliates. All rights reserved.