Log In
Or create an account ->
Imperial Library
Home
About
News
Upload
Forum
Help
Login/SignUp
Index
Preface
Who This Book Is For
O’Reilly Online Learning
How to Contact Us
Acknowledgments from the Authors
Acknowledgments from Paul Zikopoulos
1. What in the AI? How Did We Get Here?
Collecting Data in Real Time, but Understanding It in Stale Time
The Modality of Everything and the Data Collection Curve
Even Steeper: The Future of the Data Collection Curve
Where We Are Now—Haystacks, Needles, and More Data
How to Displace Today’s Disruptors
Let’s Get Ready for a Climb!
2. The Journey to AI
What Is Artificial Intelligence, Anyway?
Types of AI
Data
Models
Where AI Has Been
What Does AI Mean for Business?
The Journey to AI
All Radically New Technologies Face Resistance
Where Are We Now? And Where Are We Going?
Moving Forward
3. How to Overcome AI Failures and Challenges
AI’s Emergence in Business Today
Data
Computing Power
Investment
Early Examples of AI Success
Example: Vodafone’s TOBi Transforms the Customer Experience
Example: How a French Bank Built on Its Strength of Quality Customer Service
Early AI Failures
AI Challenges: Data, Talent, Trust
AI Challenge: Data
AI Challenge: Talent
AI Challenge: Trust
Overcoming Challenges with Advanced Research and Products
Overcoming Challenges with the Right Partner
4. The AI Ladder: A Path to Organizational Transformation
Suitability of AI
Determining the Right Business Problems to Solve with AI
Building a Data Team
Putting the Budget in Place
Developing an Approach
There Is No AI Without IA
The AI Ladder
Collect
Organize
Analyze
Infuse
Simplify, Automate, and Transform
5. Modernize Your Information Architecture
A Modern Infrastructure for AI
All Parts Are Visible
Legacy Systems Are Made Accessible or Eliminated
All Parts of the System Are Continuously Monitored
Inefficiencies Are Identified and Removed
New Architectures for IT
Data: The Fuel; Cloud: The Means
To the Cloud, and Beyond: Cloud as Capability
Fuel for the Fire
From Databases to Data Warehouses, Data Marts, and Data Lakes
Example: Wireless Carrier Architects a Solution Using Both a Data Lake and a Data Warehouse
Data Virtualization
Unifying Access to Data Through Big SQL
Object Storage as the Preferred Fabric
Open Data Stores and Open Data Formats
Next-Generation Databases
The Power of an AI Database
Streaming Data
Get the Right Tools
The Importance of Open Source Technologies
Community Thinking and Culture
High Code and Component Quality
Real Examples of Modernizing IT Infrastructure
Example: Siemens Looks to the Cloud to Unify Its Data Processes
Example: Fannie Mae Transforms with a Governed and Centralized Data Environment
Don’t Neglect the Foundation!
6. Collect Your Data
What Needs to Happen on the Collect Rung
Example: EMC Develops a Data Collection Strategy
Start with a Data Census: Learn What’s Out There
Understand Data in a Business Context, and Partner with SMEs
Getting Beyond Transactional Data
The Challenges of Collecting New Sources of High-Volume Unstructured Data
Organizational Aspects of Data Access
Example: Procter & Gamble Avoid Data Silos Using a Central Data Warehouse
Example: eBay Eliminates Data Silos by Publishing Business Processes as APIs
Ownership, Stewardship, Regulatory Compliance, and Discipline
Example: Owens-Illinois
Collecting Data: You Can Win This Battle!
7. Organize Your Data
Poor Data Leads to Poor AI
Regulation Demands Quality Data
What Needs to Happen on the Organize Rung
Cleaning Data
Documenting and Cataloging Data
Understanding Data: The “Seller” Gong Show
Metadata for Models
Maintaining the Catalog
Governing Data
Enterprise Performance Management
Example: ANZ Banking Group Embeds Sound Data Management and Governance Policies
DataOps
Now That Your Data Is Trustworthy, on to Analysis!
8. Analyze Your Data
Why Organizations Need an End-to-End AI Lifecycle
Build
Example: Using Machine Learning, an Insurer Cuts Costs and Boosts Productivity
Run
Manage
Aligning Model Output with Business Metrics
Learning, Iterating, Learning
Example: Risk Management Company Gets Creative to Offset the Expense of Training Models
Automating the AI Lifecycle
AutoAI
NeuNetS
Incorporating AI into DevOps Processes
Emphasizing Trust and Transparency
Example: By Shining Light on Data Attributes, a Bank’s AI System Demonstrates Integrity, Fairness, Explainability, and Resiliency
Example: Avoiding the “Black Box” Dilemma
Avoiding the Piecemeal Approach
Example: SaaS Company Gleans New Insights by Applying AI to Historical Data
Ready to Infuse...
9. Infuse AI Throughout the Business
Customer Service
Financial Operations
Risk and Compliance
IT Operations
Business Operations
Themes Across All Intelligent Workflows
Building the Next-Generation C-Suite
10. Tips and Best Practices on How to Get Started
Manage Organization-Wide Change
Change in Daily Tasks
Change in Overall Business Processes
Change in Thinking About Data
Make Data a Team Sport (And Some Cool History About Car Racing)
Subject Matter Experts
Data Scientists
Data Operations (DataOps) Specialists
Data Engineers
Training for Career Development
Embrace AI Centers of Excellence
Example: Honda Sets Up Knowledge Hubs to Build Minimum Viable Products, Organize Training, Share Data
Build Ethics Into Your Process
Privacy
Safety
Fairness
Building Trust in AI
Choose Projects Selectively, and Embrace Failure
Example: Insurer Tracks Metrics to Communicate Success of Its Model
Beware of False Negatives
11. The Future of AI
AI Themes to Take Us Through the Next Five Years
Theme #1: AI Is Not a Fad
Theme #2: Data-Generating Sensors Will Proliferate
Theme #3: Data Will Be Processed at the Edge
Theme #4: AI Will Spread Everywhere
Theme #5: AI Will Disappear into the Background and Become Boring
Future AI Use Cases for Business
Cybersecurity
Autonomous Driving, Autonomous Everything
Conversational Digital Agents and Personal Assistants
Real Estate
Retail
Insurance
Customer Service
The Future of Work in an AI-Driven World
A Deeper Dive into AI and Edge Computing
Using the Edge and AI for Good
Conclusion
Index
← Prev
Back
Next →
← Prev
Back
Next →