Methods & notes
Writing
The methodologies I've developed building AI systems, written up in depth: how I think about scope, correctness, and building where mistakes are expensive.
- Read →
Exact over approximate
Why I default to FAISS IndexFlatL2 instead of IndexIVFFlat for most production RAG.
- Read →
Silent failures
The most dangerous bugs in AI systems are the ones that return a confident, wrong answer.
- Read →
Scope before you build
How tightly framing the problem decides whether an AI project ships or stalls.
- Read →
Compliance first
Designing for auditability and regulation from day one, not as a retrofit.