Engineering AI Systems - Ingo Weber

. .

Résumé :

Master the Engineering of AI Systems: The Essential Guide for Architects and Developers

In today's rapidly evolving world, integrating artificial intelligence (AI) into your systems is no longer optional. Engineering AI Systems: Architecture and DevOps Essentials is a comprehensive guide to mastering the complexities of AI systems engineering. This book combines robust software architecture with cutting-edge DevOps practices to deliver high-quality, reliable, and scalable AI solutions.

Experts Len Bass, Qinghua Lu, Ingo Weber, and Liming Zhu demystify the complexities of engineering AI systems, providing practical strategies and tools for seamlessly incorporating AI in your systems. You will gain a comprehensive understanding of the fundamentals of AI and software engineering and how to combine them to create powerful AI systems. Through real-world case studies, the authors illustrate practical applications and successful implementations of AI in small- to medium-sized enterprises across various industries, and offer actionable strategies for designing, building, and operating AI systems that deliver real business value.

• Lifecycle management of AI models, from data preparation to deployment ...

  Biographie:
  

  Preface xiii
  Acknowledgments xvii
  About the Authors xix

  Chapter 1: Introduction 1
  1.1 What We Talk about When We Talk about Things: Terminology 2
  1.2 Achieving System Qualities 4
  1.3 Life-Cycle Processes 6
  1.4 Software Architecture 10
  1.5 AI Model Quality 13
  1.6 Dealing with Uncertainty 19
  1.7 Summary 20
  1.8 Discussion Questions 21
  1.9 For Further Reading 21

  Chapter 2: Software Engineering Background 23
  2.1 Distributed Computing 23
  2.2 DevOps Background 35
  2.3 MLOps Background 42
  2.4 Summary 44
  2.5 Discussion Questions 45
  2.6 For Further Reading 45

  Chapter 3: AI Background 47
  3.1 Terminology 48
  3.2 Selecting a Model 49
  3.3 Preparing the Model for Training 65
  3.4 Summary 69
  3.5 Discussion Questions 69
  3.6 For Further Reading 69

  Chapter 4: Foundation Models 71
  4.1 Foundation Models 71
  4.2 Transformer Architecture 72
  4.3 Alternatives in FM Architectures 74
  4.4 Customizing FMs 75
  4.5 Designing a System Using FMs 86
  4.6 Maturity of FMs and Organizations 91
  4.7 Challenges of FMs 93
  4.8 Summary 94
  4.9 Discussion Questions 94
  4.10 For Further Reading 94

  Chapter 5: AI Model Life Cycle 97
  5.1 Developing the Model 97
  5.2 Building the Model 108
  5.3 Testing the Model 109
  5.4 Release 114
  5.5 Summary 114
  5.6 Discussion Questions 115
  5.7 For Further Reading 115

  Chapter 6: System Life Cycle 117
  6.1 Design 118
  6.2 Developing Non-AI Modules 121
  6.3 Build 122
  6.4 Test 123
  6.5 Release and Deploy 125
  6.6 Operate, Monitor, and Analyze 135
  6.7 Summary 140
  6.8 Discussion Questions 141
  6.9 For Further Reading 141

  Chapter 7: Reliability 143
  7.1 Fundamental Concepts 143
  7.2 Preventing Faults 145
  7.3 Detecting Faults 149
  7.4 Recovering from Faults 152
  7.5 Summary 154
  7.6 Discussion Questions 154
  7.7 For Further Reading 154

  Chapter 8: Performance 155
  8.1 Efficiency 155
  8.2 Accuracy 164
  8.3 Summary 173
  8.4 Discussion Questions 173
  8.5 For Further Reading 174

  Chapter 9: Security 175
  9.1 Fundamental Concepts 176
  9.2 Approaches to Mitigating Security Concerns 180
  9.3 Summary 188
  9.4 Discussion Questions 189
  9.5 For Further Reading 189

  Chapter 10: Privacy and Fairness 191
  10.1 Privacy in AI Systems 192
  10.2 Fairness in AI Systems 193
  10.3 Achieving Privacy 194
  10.4 Achieving Fairness 197
  10.5 Summary 201
  10.6 Discussion Questions 201
  10.7 For Further Reading 202

  Chapter 11: Observability 203
  11.1 Fundamental Concepts 203
  11.2 Evolving from Monitorability to Observability 204
  11.3 Approaches for Enhancing Observability 207
  11.4 Summary 211
  11.5 Discussion Questions 211
  11.6 For Further Reading 212

  Chapter 12: The Fraunhofer Case Study: Using a Pretrained Language Model for Tendering 213
  12.1 The Problem Context 214
  12.2 Case Study Description and Setup 217
  12.3 Summary 232
  12.4 Takeaways 233
  12.5 Discussion Questions 233
  12.6 For Further Reading 233

  Chapter 13: The ARM Hub Case Study: Chatbots for Small and Medium-Size Australian Enterprises 235
  13.1 Introduction 235
  13.2 Our Approach 236
  13.3 LLMs in SME Manufacturing 238
  13.4 A RAG-Based Chatbot for SME Manufacturing 238
  13.5 Architecture of the ARM Hub Chatbot 239
  13.6 MLOps in ARM Hub 244
  13.7 Ongoing Work 251
  1...

  Sommaire:
  

• Best practices in system architecture and DevOps for AI systems
• System reliability, performance, and security in AI implementations
• Privacy and fairness in AI systems to build trust and achieve compliance
• Effective monitoring and observability for AI systems to maintain operational excellence
• Future trends in AI engineering to stay ahead of the curve

Equip yourself with the tools and understanding to lead your organization's AI initiatives. Whether you are a technical lead, software engineer, or business strategist, this book provides the essential insights you need to successfully engineer AI systems.

Register your book for convenient access to downloads, updates, and/or corrections as they become available. See inside book for details.