About the Book

“Patterns of Application Development Using AI” is a groundbreaking book that explores the intersection of artificial intelligence (AI) and application development. In this book, Obie Fernandez, a renowned software developer and co-founder of AI-powered consultant platform Olympia, shares his invaluable insights and experiences from a year-long journey of building an AI-powered application.

Through a compelling combination of narrative chapters and practical pattern references, Obie presents a comprehensive guide to leveraging the power of large language models (LLMs) in application development. He introduces innovative patterns such as the “Multitude of Workers,” “Self-Healing Data,” and “Contextual Content Generation,” which empower developers to build intelligent, adaptive, and user-centric applications.

Unlike other books on AI that focus on theoretical concepts or delve into the intricacies of machine learning algorithms, this book takes a pragmatic approach. It provides concrete examples, real-world use cases, and actionable advice on how to integrate AI components and functions into application architectures. Obie shares his successes, challenges, and lessons learned, offering a unique perspective on the practical application of AI in software development.

Table of Contents

1. Foreword by Gregor Hohpe
2. Preface
3. About the Book
4. About the Code Examples
5. What I Don’t Cover
6. Who This Book Is For
7. Building a Common Vocabulary
8. Getting Involved
9. Acknowledgments
10. What’s with the illustrations?
11. About Lean Publishing
12. About The Author
13. Introduction
14. Thoughts on Software Architecture
15. What is a Large Language Model?
16. Understanding Inference
17. Thinking About Performance
18. Experimenting With Different LLM Models
19. Compound AI Systems
20. Part 1: Fundamental Approaches & Techniques
21. Narrow The Path
22. Latent Space: Incomprehensibly Vast
23. How The Path Gets “Narrowed”
24. Raw Versus Instruct-Tuned Models
25. Prompt Engineering
26. Prompt Distillation
27. What about fine-tuning?
28. Retrieval Augmented Generation (RAG)
29. What is Retrieval Augmented Generation?
30. How Does RAG Work?
31. Why Use RAG in Your Applications?
32. Implementing RAG in Your Application
33. Proposition Chunking
34. Real-World Examples of RAG
35. Intelligent Query Optimization (IQO)
36. Reranking
37. RAG Assessment (RAGAs)
38. Challenges and Future Outlook
39. Multitude of Workers
40. AI Workers As Independent Reusable Components
41. Account Management
42. E-commerce Applications
43. Healthcare Applications
44. AI Worker as a Process Manager
45. Integrating AI Workers Into Your Application Architecture
46. Composability and Orchestration of AI Workers
47. Combining Traditional NLP with LLMs
48. Tool Use
49. What is Tool Use?
50. The Potential of Tool Use
51. The Tool Use Workflow
52. Best Practices for Tool Use
53. Composing and Chaining Tools
54. Future Directions
55. Stream Processing
56. Implementing a ReplyStream
57. The “Conversation Loop”
58. Auto Continuation
59. Conclusion
60. Self Healing Data
61. Practical Case Study: Fixing Broken JSON
62. Considerations and Counterindications
63. Contextual Content Generation
64. Personalization
65. Productivity
66. Rapid Iteration and Experimentation
67. AI Powered Localization
68. The Importance of User Testing and Feedback
69. Generative UI
70. Generating Copy for User Interfaces
71. Defining Generative UI
72. Example
73. The Shift to Outcome-Oriented Design
74. Challenges and Considerations
75. Future Outlook and Opportunities
76. Intelligent Workflow Orchestration
77. Business Need
78. Key Benefits
79. Key Patterns
80. Exception Handling and Recovery
81. Implementing Intelligent Workflow Orchestration in Practice
82. Monitoring and Logging
83. Scalability and Performance Considerations
84. Testing and Validation of Workflows
85. Part 2: The Patterns
86. Prompt Engineering
87. Chain of Thought
88. Mode Switch
89. Role Assignment
90. Prompt Object
91. Prompt Template
92. Structured IO
93. Prompt Chaining
94. Prompt Rewriter
95. Response Fencing
96. Query Analyzer
97. Query Rewriter
98. Ventriloquist
99. Discrete Components
100. Predicate
101. API Facade
102. Result Interpreter
103. Virtual Machine
104. Human In The Loop (HITL)
105. High-Level Patterns
106. Escalation
107. Feedback Loop
108. Passive Information Radiation
109. Collaborative Decision Making (CDM)
110. Continuous Learning
111. Ethical Considerations
112. Technological Advancements and Future Outlook
113. Intelligent Error Handling
114. Traditional Error Handling Approaches
115. Contextual Error Diagnosis
116. Intelligent Error Reporting
117. Predictive Error Prevention
118. Smart Error Recovery
119. Personalized Error Communication
120. Adaptive Error Handling Workflow
121. Quality Control
122. Eval
123. Guardrail
124. Guardrails and Evals: Two Sides of the Same Coin
125. Glossary
126. Notes