Concept map: RAG pipeline

Taught and used in

RAG and retrieval

Insights 48

• Score embeddings with the similarity function the model was trained for
• Re-embed the whole corpus whenever the embedding model or its preprocessing changes
• Test effective context at your own input lengths instead of trusting the advertised window
• Do not fine-tune to inject facts that change; retrieve them instead
• Chunk along the document's own structure before tuning chunk size
• Give each chunk the context it lost when it was cut out of the document
• Compare chunk sizes at the same context token budget, not the same k
• Enforce validity dates and scope with metadata filters, not with the prompt

Projects 12

• Flagship: passenger-rights document Q&A with a retrieval baseline, one controlled improvement and an honest generation mode
• Hybrid lexical/latent retrieval and lightweight reranking ablation over the academy's own knowledge base
• Chunking experiment on long technical papers: evidence recall at a fixed context budget
• Tool-calling agent vs deterministic workflow for rebooking proposals after a cancellation
• Prompt-injection red-team harness for a product-Q&A assistant, run on safe fixtures
• LLM-as-judge reliability study: agreement with human pairwise labels, position bias and length bias