Python & AI Systems Engineering
RAG & Document Intelligence
Retrieval systems built as measurable pipelines rather than opaque demos.
Ingestion, parsing, chunking, embeddings, indexing, retrieval, and reranking
Hybrid retrieval with FAISS/pgvector and optional cross-encoder reranking
Evaluation, retrieval telemetry, index health, and failure diagnosis
When I'm a good fit
Relevant proof includes DocIntel/RAG Intelligence: an end-to-end retrieval platform with pipeline-level observability and evaluation.
I focus on production behavior: explicit interfaces, state and persistence, bounded model calls, reliability, cost, and deployability. I am usually not the best fit for frontend-only work or isolated prompt-writing tasks.