Lead Data Scientist
4 ratings and 11 bookings across all services

For engineers building a RAG system at work or for a serious portfolio project.
We'll go through your pipeline (chunking, embeddings, retrieval, reranking, prompts, evaluation) and find what's limiting it: wrong answers, slow responses, or cost per query. You leave knowing what to change first, and how to measure whether it worked.
After the call: a short written list of changes, in priority order.
Before the call: what the system does, your stack, the main problem (accuracy, speed or cost), and a repo link or architecture diagram if you can share one. Share only what you're allowed to: no confidential company code.