AI Dev Landscape¶
The AI developer ecosystem moves weekly. This is my map — what exists, how it works, and what's worth adopting.
What This Is¶
A personal field guide to the AI tooling landscape. Not a tutorial collection or link dump — a structured reference for understanding how these tools work, how they relate, and what's worth investing time in.
The Stack¶
The AI developer ecosystem has layers, and understanding where a tool sits changes how you evaluate it:
| Layer | What It Does | Key Tools |
|---|---|---|
| Fundamentals | How models are trained | Transformers, RLHF, DPO, distillation |
| Inference | Running models | Ollama, vLLM, llama.cpp, SGLang |
| Coding Assistants | AI writes code | Claude Code, Codex, Cursor, Copilot |
| Agent Frameworks | Autonomous tool use | OpenClaw, LangGraph, CrewAI, smolagents |
| Protocols | Integration standards | MCP, tool use, A2A |
| RAG | Knowledge retrieval | LightRAG, LlamaIndex, Chroma, Qdrant |
| Patterns | How to build with AI | Skills, hooks, agentic workflows |
How to Use This Site¶
- Looking up a specific tool? Check the relevant section in the sidebar.
- Want depth? Deep reads have detailed analyses of key tools.
- Tracking trends? The Trends section is updated monthly.
- Comparing options? Category overview pages have comparison tables.
Quick Links¶
- Claude Code Deep Dive — how skills, hooks, and MCP work together
- MCP Protocol — the standard for tool integration
- Agent Framework Landscape — what exists and how they differ
- Local Inference — running models on your Mac