Technical authority
Expert explanations for hard RAG decisions
Curated authority pages built from indexed engineering talks — best explanations, comparisons, failure modes, and when-to-use guides. No keyword spam; every clip links to a real timestamp.
Best explanations
Foundational walkthroughs — what RAG is and how each layer works.
Comparisons
X vs Y tradeoffs with decision rules and expert clips.
- RAG vs fine-tuning — when to use each
- RAG vs semantic search — retrieval-only vs grounded generation
- RAG vs MCP — context retrieval vs tool protocol
- RAG vs AI agents — knowledge grounding vs action planning
- RAG vs agentic RAG — static retrieval vs orchestrated retrieval loops
- RAG evaluation vs observability — offline benchmarks vs production traces
- LangChain vs LlamaIndex — orchestration vs indexing workflow
- Pinecone vs Weaviate — managed vector infra vs hybrid database tradeoffs
- Chunking vs reranking — index boundaries vs post-retrieval ordering
Failure modes
How retrieval and grounding fail before you tune prompts.
When to use
When to choose one architecture path over another.
Registry: lib/authority/authority-page-registry.ts