Enterprise AI in the Cloud: A Practical Guide to Deploying End-to-End Machine Learning and ChatGPT Solutions

Cover....1 Title Page....5 Copyright Page....6 Acknowledgments....9 About the Author....11 About the Technical Editor....13 Contents....15 Introduction....19 How This Book Is Organized....19 Who Should Read This Book?....20 Data Scientists and AI Teams....20 IT Leaders and Teams....20 Students and Academia....21 Consultants and Advisors....21 Business Strategists and Leaders....21 C-Level Executives....21 Why You Should Read This Book....21 Unique Features....21 Comprehensive Coverage of All Aspects of Enterprise-wide AI Transformation....22 Case Study Approach....22 Coverage of All Major Cloud Platforms....22 Discussion of Nontechnical Aspects of AI....22 Best Practices for MLOps and AI Governance....22 Up-to-Date Content....22 Hands-on Approach....22 Part I Introduction....23 Chapter 1 Enterprise Transformation with AI in the Cloud....25 Understanding Enterprise AI Transformation....25 Why Some Companies Succeed at Implementing AI and ML While Others Fail....26 Transform Your Company by Integrating AI, ML and Gen AI into Your Business Processes....26 Adopt AI-First to Become World-Class....27 Importance of an AI-First Strategy....27 Prioritize AI and Data Initiatives....27 Leveraging Enterprise AI Opportunities....28 Enable One-to-One, Personalized, Real-Time Service for Customers at Scale....28 Enterprise-wide AI Opportunities....30 Growing Industry Adoption of AI....33 Workbook Template - Enterprise AI Transformation Checklist....37 Summary....38 Review Questions....38 Answer Key....40 Chapter 2 Case Studies of Enterprise AI in the Cloud....41 Case Study 1: The U.S. Government and the Power of Humans and Machines Working Together to Solve Problems at Scale....41 Revolutionizing Operations Management with AI/ML....43 Enabling Solutions for Improved Operations....43 Case Study 2: Capital One and How It Became a Leading Technology Organization in a Highly Regulated Environment....43 Building Amazing Experiences Due to Data Consolidation....44 Becoming Agile and Scalable by Moving Data Centers Into the Cloud....44 Building a Resilient System by Embracing Cloud-Native Principles....45 Impact of Cloud-First Thinking on DevOps, Agile Development, and Machine Learning....45 Becoming an AI-First Company: From Cloud Adoption to Thrilling Customer Experiences....45 Case Study 3: Netflix and the Path Companies Take to Become World-Class....46 Cloud and AI Technology: A Game-Changer for Netflix’s Business Model and Success....46 Cloud Infrastructure and AI Adoption Drives Process Transformation....47 Process Transformation Drives Organizational Change....48 Workbook Template - AI Case Study....49 Summary....49 Review Questions....49 Answer Key....50 Part II Strategizing and Assessing for AI....51 Chapter 3 Addressing the Challenges with Enterprise AI....53 Challenges Faced by Companies Implementing Enterprise-wide AI....53 Business-Related Challenges....54 Data- and Model-Related Challenges....55 Platform-Related Challenges....55 How Digital Natives Tackle AI Adoption....57 They Are Willing to Take Risks....57 They Have an Advantage in Data Collection and Curation Capabilities....57 They Attract Top Talent Through Competitive Compensation and Perks....57 Get Ready: AI Transformation Is More Challenging Than Digital Transformation....57 Complexities of Skill Sets, Technology, and Infrastructure Integration....57 The Importance of Data Infrastructure and Governance....58 Change Management to Redefine Work Processes and Employee Mindsets....58 Regulatory Concerns: Addressing Bias, Ethical, Privacy, and Accountability Risks....59 Choosing Between Smaller PoC Point Solutions and Large-Scale AI Initiatives....59 The Challenges of Implementing a Large-Scale AI Initiative....59 Navigate the Moving Parts, Stakeholders, and Technical Infrastructure....59 Resource Allocation Challenges in Large-Scale AI Initiatives....59 Overcome Resistance to Change....59 Data Security, Privacy, Ethics, Compliance, and Reputation....60 Build a Business Case for Large-Scale AI Initiatives....60 Factors to Consider....60 Workbook Template: AI Challenges Assessment....61 Summary....61 Review Questions....61 Answer Key....62 Chapter 4 Designing AI Systems Responsibly....63 The Pillars of Responsible AI....63 Robust AI....65 Collaborative AI....65 Trustworthy AI....66 Scalable AI....66 Human-centric AI....67 Workbook Template: Responsible AI Design Template....70 Summary....70 Review Questions....70 Answer Key....71 Chapter 5 Envisioning and Aligning Your AI Strategy....72 Step-by-Step Methodology for Enterprise-wide AI....72 The Envision Phase....73 The Align Phase....75 Workbook Template: Vision Alignment Worksheet....77 Summary....77 Review Questions....78 Answer Key....78 Chapter 6 Developing an AI Strategy and Portfolio....79 Leveraging Your Organizational Capabilities for Competitive Advantage....79 Focus Areas to Build Your Competitive Advantage....80 Driving Competitive Advantage Through AI....81 Initiating Your Strategy and Plan to Kickstart Enterprise AI....81 Manage Your AI Strategy, Portfolio, Innovation, Product Lifecycle, and Partnerships....82 Define Your AI Strategy to Achieve Business Outcomes....82 Prioritize Your Portfolio....84 Strategy and Execution Across Phases....85 Workbook Template: Business Case and AI Strategy....87 Summary....87 Review Questions....87 Answer Key....87 Chapter 7 Managing Strategic Change....88 Accelerating Your AI Adoption with Strategic Change Management....89 Phase 1: Develop an AI Acceleration Charter and Governance Mechanisms for Your AI Initiative....89 Phase 2: Ensure Leadership Alignment....91 Phase 3: Create a Change Acceleration Strategy....94 Workbook Template: Strategic Change Management Plan....97 Summary....97 Review Questions....97 Answer Key....98 Part III Planning and Launching a Pilot Project....99 Chapter 8 Identifying Use Cases for Your AI/ML Project....101 The Use Case Identification Process Flow....102 Educate Everyone as to How AI/ML Can Solve Business Problems....102 Define Your Business Objectives....103 Identify the Pain Points....103 Start with Root-Cause Analysis....104 Identify the Success Metrics....105 Explore the Latest Industry Trends....106 Review AI Applications in Various Industries....106 Map the Use Case to the Business Problem....109 Prioritizing Your Use Cases....109 Define the Impact Criteria....109 Define the Feasibility Criteria....109 Assess the Impact....110 Assess the Feasibility....110 Prioritize the Use Cases....110 Review and Refine the Criteria....111 Choose the Right Model....111 Use Cases to Choose From....113 AI Use Cases for DevOps....114 AI for Healthcare and Life Sciences....114 AI Enabled Contact Center Use Cases....114 Business Metrics Analysis....114 Content Moderation....115 AI for Financial Services....115 Cybersecurity....116 Digital Twinning....116 Identity Verification....117 Intelligent Document Processing....117 Intelligent Search....117 Machine Translation....118 Media Intelligence....119 ML Modernization....119 ML-Powered Personalization....120 Computer Vision....120 Personal Protective Equipment....121 Generative AI....121 Workbook Template: Use Case Identification Sheet....126 Summary....126 Review Questions....126 Answer Key....127 Chapter 9 Evaluating AI/ML Platforms and Services....128 Benefits and Factors to Consider When Choosing an AI/ML Service....129 Benefits of Using Cloud AI/ML Services....129 Factors to Consider When Choosing an AI/ML Service....131 AWS AI and ML Services....134 AI Services....134 Amazon SageMaker....134 AI Frameworks....134 Differences Between Machine Learning Algorithms, Models, and Services....135 Core AI Services....135 Text and Document Services....136 Chatbots: Amazon Lex....138 Speech....139 Vision Services....140 Specialized AI Services....143 Business Processing Services....143 Kendra for Search....147 Code and DevOps....148 Industrial Solutions....151 Healthcare Solutions....152 Machine Learning Services....156 Amazon SageMaker....156 Amazon SageMaker Canvas....157 SageMaker Studio Lab....157 The Google AI/ML Services Stack....158 For Data Scientists....158 For Developers....160 The Microsoft AI/ ML Services Stack....164 Azure Applied AI Services....164 Azure Cognitive Services....164 Azure Machine Learning....167 Other Enterprise Cloud AI Platforms....169 Dataiku....169 DataRobot....169 KNIME....169 IBM Watson....169 Salesforce Einstein AI....169 Oracle Cloud AI....169 Workbook Template: AI/ML Platform Evaluation Sheet....170 Summary....170 Review Questions....171 Answer Key....173 Chapter 10 Launching Your Pilot Project....174 Launching Your Pilot....175 Planning for Launch....175 Recap of the Envision Phase....175 Planning for the Machine Learning Project....176 Following the Machine Learning Lifecycle....177 Business Goal Identification....177 Machine Learning Problem Framing....178 Data Processing....178 Model Development....178 Model Deployment....179 Model Monitoring....179 Workbook Template: AI/ML Pilot Launch Checklist....180 Summary....181 Review Questions....181 Answer Key....181 Part IV Building and Governing Your Team....183 Chapter 11 Empowering Your People Through Org Change Management....185 Succeeding Through a People-centric Approach....186 Evolve Your Culture for AI Adoption, Innovation, and Change....188 Redesign Your Organization for Agility and Innovation with AI....190 Aligning Your Organization Around AI Adoption to Achieve Business Outcomes....190 Workbook Template: Org Change Management Plan....192 Summary....193 Review Questions....193 Answer Key....194 Note....194 Chapter 12 Building Your Team....195 Understanding the Roles and Responsibilities in an ML Project....195 Build a Cross-Functional Team for AI Transformation....195 Adopt Cloud and AI to Transform Current Roles....196 Customize Roles to Suit Your Business Goals and Needs....196 Workbook Template: Team Building Matrix....204 Summary....204 Review Questions....204 Answer Key....205 Part V Setting Up Infrastructure and Managing Operations....207 Chapter 13 Setting Up an Enterprise AI Cloud Platform Infrastructure....209 Reference Architecture Patterns for Typical Use Cases....210 Customer 360-Degree Architecture....210 Develop an Event-Driven Architecture Using IoT Data....213 Personalized Recommendation Architecture....215 Real-Time Customer Engagement....217 Data Anomaly and Fraud Detection....220 Factors to Consider When Building an ML Platform....222 The Build vs. Buy Decision....222 Choosing Between Cloud Providers....226 Key Components of an ML and DL Platform....228 Key Components of an Enterprise AI/ML Healthcare Platform....228 Data Management Architecture....229 Data Science Experimentation Platform....231 Hybrid and Edge Computing....233 The Multicloud Architecture....235 Workbook Template: Enterprise AI Cloud Platform Setup Checklist....236 Summary....236 Review Questions....237 Answer Key....238 Chapter 14 Operating Your AI Platform with MLOps Best Practices....239 Central Role of MLOps in Bridging Infrastructure, Data, and Models....239 What Is MLOps?....239 Automation Through MLOps Workflows....240 Model Operationalization....243 Automation Pipelines....244 Deployment Scenarios....247 Model Inventory Management....247 Logging and Auditing....250 Data and Artifacts Lineage Tracking....251 Container Image Management....254 Tag Management....255 Workbook Template: ML Operations Automation Guide....259 Summary....259 Review Questions....259 Answer Key....261 Part VI Processing Data and Modeling....263 Chapter 15 Process Data and Engineer Features in the Cloud....265 Understanding Your Data Needs....266 Benefits and Challenges of Cloud-Based Data Processing....269 Benefits of Cloud-Based Data Processing....269 Challenges of Cloud-Based Data Processing....269 Handling Different Types of Data....269 The Data Processing Phases of the ML Lifecycle....272 Data Collection and Ingestion....272 Data Storage Options....274 Understanding the Data Exploration and Preprocessing Stage....275 Data Preparation....275 Data Preprocessing....276 Feature Engineering....281 Feature Types....281 Feature Selection....282 Feature Extraction....283 Feature Creation....284 Feature Transformation....285 Feature Imputation....286 Workbook Template: Data Processing & Feature Engineering Workflow....287 Summary....287 Review Questions....287 Answer Key....289 Chapter 16 Choosing Your AI/ML Algorithms....290 Back to the Basics: What Is Artificial Intelligence?....291 Machine Learning: The Brain Behind Artificial Intelligence....291 Features and Weights of Predictive Algorithms....292 Factors to Consider When Choosing a Machine Learning Algorithm....292 Data-Driven Predictions Using Machine Learning....294 Different Categories of Machine Learning....295 Using Supervised Learning....296 Types of Supervised Learning Algorithms....298 Using Unsupervised Learning to Discover Patterns in Unlabeled Data....314 Reinforced Learning: Learning by Trial and Error....322 Deep Learning....324 Convolutional Neural Networks....325 Recurrent Neural Networks....327 Transformer Models....328 Generative Adversarial Networks....330 The AI/ML Framework....331 TensorFlow and PyTorch....331 Keras....332 Caffe....332 MXNet....333 Scikit....333 Chainer....333 Workbook Template: AI/ML Algorithm Selection Guide....333 Summary....333 Review Questions....334 Answer Key....336 Chapter 17 Training, Tuning, and Evaluating Models....337 Model Building....337 Structure, Parameters, and Hyperparameters....338 Steps Involved During Model Building....339 Model Training....340 Distributed Training....340 Problems Faced When Training Models....341 Training Code Container....342 Model Artifacts....343 Model Tuning....344 Hyperparameters....344 Choosing the Right Hyperparameter Optimization Technique....346 Model Validation....347 Choosing the Right Validation Techniques....347 Validation Metrics....348 Validation Metrics for Classification Problems....349 Model Evaluation....352 Best Practices....352 Streamlining Your ML Workflows Using MLOps....353 Securing Your ML Platform....354 Building Robust and Trustworthy Models....356 Ensuring Optimal Performance and Efficiency....357 Utilizing Cost Optimization Best Practices....357 Workbook Template: Model Training and Evaluation Sheet....360 Summary....361 Review Questions....361 Answer Key....363 Part VII Deploying and Monitoring Models....365 Chapter 18 Deploying Your Models Into Production....367 Standardizing Model Deployment, Monitoring, and Governance....367 Challenges in Model Deployment Monitoring and Governance....368 Deploying Your Models....369 Pre-deployment Checklist....369 Deployment Process Checklist....370 Choosing the Right Deployment Option....370 Choosing an Appropriate Deployment Strategy....372 Choosing Between Real-Time and Batch Inference....374 Implementing an Inference Pipeline....375 Synchronizing Architecture and Configuration Across Environments....377 Ensuring Consistency in the Architecture....378 Ensuring Identical Performance Across Training and Production....378 Looking for Bias in Training and Production....378 Generating Governance Reports....378 MLOps Automation: Implementing CI/CD for Models....379 Workbook Template: Model Deployment Plan....381 Summary....381 Review Questions....381 Answer Key....382 Chapter 19 Monitoring Models....383 Monitoring Models....384 Importance of Monitoring Models in Production....384 Challenges Faced When Monitoring Models....384 Key Strategies for Monitoring ML Models....385 Detecting and Addressing Data Drift....385 Detecting and Addressing Concept Drift....386 Monitoring Bias Drift....387 Watching for Feature Attribution Drift....387 Model Explainability....388 Tracking Key Model Performance Metrics....388 Classification Metrics....388 Regression Metrics....388 Clustering Metrics....389 Ranking Metrics....389 Real-Time vs. Batch Monitoring....389 When to Use Real-Time Monitoring....389 When to Use Batch Monitoring....389 Tools for Monitoring Models....389 Cloud Provider Tools....390 Open-Source Libraries....390 Third-Party Tools....390 Building a Model Monitoring System....390 Determining the Model Metrics to be Monitored....391 Setting Up the Thresholds for Monitoring....391 Employing a Monitoring Service with Dashboards....391 Setting Up Alerts....391 Conducting Periodic Reviews....391 Monitoring Model Endpoints....392 Automating Endpoint Changes Through a Pipeline....392 Implementing a Recoverable Endpoint....392 Implementing Autoscaling for the Model Endpoint....393 Optimizing Model Performance....395 Reviewing Features Periodically....396 Implementing a Model Update Pipeline....396 Keeping Models Fresh with a Scheduler Pipeline....396 Workbook Template: Model Monitoring Tracking Sheet....397 Summary....397 Review Questions....397 Answer Key....398 Chapter 20 Governing Models for Bias and Ethics....399 Importance of Model Governance....400 Strategies for Fairness....400 Addressing Fairness and Bias in Models....401 Addressing Model Explainability and Interpretability....401 Ethical Considerations for Deploying Models....402 Implementing Augmented AI for Human Review....403 Operationalizing Governance....403 Tracking Your Models....403 Managing Model Artifacts....405 Controlling Your Model Costs Using Tagging....407 Setting Up a Model Governance Framework....407 Workbook Template: Model Governance for Bias & Ethics Checklist....410 Summary....410 Review Questions....410 Answer Key....410 Part VIII Scaling and Transforming AI....411 Chapter 21 Using the AI Maturity Framework to Transform Your Business....413 Scaling AI to Become an AI-First Company....414 Why Do You Need a Maturity Model Framework?....415 The AI Maturity Framework....416 The Five Stages of Maturity....416 The Six Dimensions of AI Maturity....421 Workbook Template: AI Maturity Assessment Tool....427 Summary....428 Review Questions....428 Answer Key....428 Chapter 22 Setting Up Your AI COE....429 Scaling AI to Become an AI-First Company....430 Establishing an AI Center of Excellence....431 From Centralized Unit to Enterprise-wide Advisor....432 Evolving from Strategy to Operations Focus....432 Workbook Template: AI Center of Excellence (AICOE) Setup Checklist....435 Summary....435 Review Questions....435 Answer Key....437 Chapter 23 Building Your AI Operating Model and Transformation Plan....438 Understanding the AI Operating Model....439 The Purpose of the AI Operating Model....439 When Do You Implement an AI Operating Model?....440 Implementing Your AI Operating Model....440 Customer-centric AI Strategy to Drive Innovation....440 Developing an AI Transformation Plan....446 Workbook Template: AI Operating Model and Transformation Plan....450 Summary....450 Review Questions....451 Answer Key....451 Part IX Evolving and Maturing AI....453 Chapter 24 Implementing Generative AI Use Cases with ChatGPT for the Enterprise....455 The Rise and Reach of Generative AI....456 The Powerful Evolution of Generative AI....456 The Power of Generative AI/ChatGPT for Business Transformation and Innovation....462 The Fascinating World of Generative AI: From GANs to Diffusion Models....465 Implementing Generative AI and ChatGPT....467 Best Practices When Implementing Generative AI and ChatGPT....470 Strategy Considerations for Generative AI and ChatGPT....471 Challenges of Generative AI and ChatGPT....472 Strategy for Managing and Mitigating Risks....474 Generative AI Cloud Platforms....475 Google’s Generative AI Cloud Platform Tools....476 AWS Generative AI Cloud Platform Tools....477 Azure Generative AI Cloud Platform Tools....479 Additional Tools and Platforms....483 Workbook Template: Generative AI Use Case Planner....485 Summary....485 Review Questions....485 Answer Key....486 Chapter 25 Planning for the Future of AI....487 Emerging AI Trends....488 Smart World....488 AR and VR Technology....488 Metaverse....489 Digital Humans and Digital Twins....489 The Productivity Revolution....491 AI in the Edge....491 Intelligent Apps....491 Compressed Models....492 Self-Supervised Learning....492 Critical Enablers....493 Foundation Models....493 Knowledge Graphs....493 Hyper-Automation....494 Democratization of AI/ML....494 Transformer Models....494 Keras and TensorFlow in the Cloud....494 Quantum Machine Learning....495 Emerging Trends in Data Management....497 Federated Learning....497 AutoML....497 Data Flywheels....497 DataOps and Data Stewardship....497 Distributed Everything....498 Workbook Template: Future of AI Roadmap....498 Summary....498 Review Questions....498 Answer Key....500 Chapter 26 Continuing Your AI Journey....501 Reflecting On Your Progress....502 Reviewing the Lessons Learned....502 Exploring Opportunities for Improvement....502 Embracing the Culture of Continuous Improvement....502 Planning for the Future: Building a Roadmap....503 Mapping Your AI/ML Opportunities....503 Prioritizing Your AI/ML Opportunities....503 Mobilizing Your Team for the Journey....503 Ensuring Responsible AI/ML implementation....504 Enabling Awareness Around AI Risks and Data Handling....504 Implementing Data Security, Privacy, and Ethical Safeguards....504 Defining Ethical Framework and Data Usage Policies....504 Preparing for the Challenges Ahead....504 Encouraging Innovation, Collaboration, and High-Performing Teams....504 Leveraging the Transformational Nature of AI....505 My Personal Invite to Connect....505 Index....507 EULA....527