Flagship · RAG + MCP

Ask about Kumail's work

Answers come only from his resume and project READMEs, with citations. Built with hybrid retrieval (BM25 + embeddings) and an MCP server.

86.4% BM25 hit@1 · 22-question eval100% Embeddings hit@135 Tests · CI green

Example questions
  1. Ask Kumail Bot

    Hi — I'm the Ask Kumail bot, here to answer questions about Kumail and his work. Every answer I give comes only from his resume and project docs, with citations you can check. Try one of the examples below, or ask your own question.

Evidence · Ask Kumail

Measured, not claimed

Hit@1 on a 22-question evaluation: did the first retrieved result contain the answer? BM25 from scratch gets most of the way there — embeddings close the gap, and fusing the two is what ships.

The same numbers, as a table
MethodHit@1Eval sizeNote
BM25 (written from scratch)86.4%22 questionsKeyword ranking, no libraries
Hybrid90.9%22 questionsReciprocal rank fusion
Embeddings100%22 questionsSemantic ranking
Reciprocal rank fusionShips in the product—Merges both rankings for the final answer

Use it from Claude or any MCP client

Add this as a remote MCP server. Tools: search_docs, ask_kumail, list_projects.

/api/mcp

Source on GitHub · Retrieval eval results · Portfolio