- Architectural Depth: Moves beyond "how to call an API" to "how to build reliable systems."
- Production-Hardened: Focused exclusively on patterns that survive 24/7 enterprise traffic.
- Vendor Agnostic: Teaches the fundamental physics of LLMs, retrieval, and agents—not just specific SDKs.
- Target Audience: Competent software engineers, SREs, and architects moving from prototypes to production systems.
- Not for: Absolute beginners seeking basic tutorials or practitioners looking only for unannotated link lists.
- Foundations of AI Engineering
- Prompt Engineering Patterns
- RAG & Retrieval Systems
- Agents, Tools & MCP
- Evals & Testing
- Production Ops & Reliability
Important
⭐ Star this repository to follow updates. We track daily shifts in the AI Engineering landscape to keep this manual at the cutting edge.
| Metric | Value |
|---|---|
| Chapters | 12 |
| Proven Patterns | 50+ |
| Annotated Papers | 30+ |
| Vetted Tools | 40+ |
| Target Reliability | 99.9% |
We welcome contributions that meet our high architectural bar. Please read our CONTRIBUTING.md and STYLE_GUIDE.md before submitting a PR.
© 2026 AI Engineer Vault. Licensed under MIT.