Retrieval-augmented generation for documentation QA under a 2,000-token context budget — multi-resolution chunking, hybrid BM25+FAISS retrieval, cross-encoder reranking, and grounded generation, with a pre-registered ablation sweep and an LLM-as-a-judge evaluation harness.
bm25 embeddings-similarity hybrid-search cross-encoder langchain-python rag-pipeline retrival-augmented-generation context-engineering fasiss
-
Updated
Jul 10, 2026 - Python