Pathan Afnan Khan · AI Engineering Field Notes
Practical notes on retrieval, agents, evaluation and observability — the parts of AI engineering that decide whether a system survives production.
AI Engineering
A practical guide to improving production RAG through chunking, metadata, hybrid retrieval, reranking and explicit evaluation instead of relying on a larger language model alone.
AI Engineering
Why production AI agents become more reliable when reasoning is paired with typed tools, validation, explicit state, bounded retries, approvals and deterministic business logic.
AI Engineering
A production-oriented observability framework for LLM systems covering traces, prompts, retrieval evidence, tool calls, latency, cost, evaluation and user feedback.