About AI Pattern Library
The AI Pattern Library is an open-source, community-driven catalog of proven AI engineering patterns. Educational content about AI patterns is scattered across blog posts, research papers, and tutorials. This project provides a structured, searchable, and executable reference.
What's a Pattern?
Each pattern in this library represents a well-established approach to solving a common AI engineering challenge. Patterns include:
- Explanation — Clear description of what the pattern does and when to use it
- Architecture Diagram — Visual representation using Mermaid diagrams
- Runnable Code — Python examples you can execute right in the browser
- Gotchas — Real-world pitfalls and how to avoid them
- Related Patterns — Links to complementary or alternative approaches
Categories
Retrieval
Patterns for retrieving and grounding with external data (RAG, hybrid search, etc.)
Reasoning
Chain-of-thought, tree-of-thought, self-consistency, and structured reasoning
Tool Use
Function calling, API integration, and tool selection strategies
Memory
Short-term, long-term, and episodic memory for conversational AI
Agents
Autonomous agents, multi-agent systems, and agent architectures
Evaluation
LLM-as-judge, automated testing, and quality assurance patterns
Orchestration
Chaining, routing, parallelization, and workflow patterns
Safety
Guardrails, content filtering, output validation, and responsible AI
Contributing
This is an open-source project and we welcome contributions! You can add new patterns, improve existing ones, fix bugs, or enhance the site.
See our Contributing Guide for details on how to submit a pattern via GitHub PR.
Tech Stack
- Astro — Static site generator
- MDX — Content with interactive components
- Mermaid — Architecture diagrams
- Pyodide — In-browser Python execution
- GitHub Pages — Hosting