Topic
Ingestion, chunking, retrieval, and grounding pipelines that feed retrieved document context to an LLM.
Lookup and aggregation are opposite retrieval problems, and chunk similarity only answers one of them. The second failure is architectural, not a tuning bug.
Reality
Lakshman Pandey's write-up maps a production retrieval system to all four NIST functions. The controls that exist are documents and evals; the ones that would gate a deploy are still marked future.
Reality
A scanned 346-page Kannada novel broke pure vector search. Hybrid BM25 plus dense retrieval with RRF, and a regex page router, took reported faithfulness to 0.92.
Reality
A dev.to write-up argues per-request capture of prompt, model version, tokens, cost, latency and output quality is now a required line item, because crash-oriented telemetry has no field for any of it.
Reality
An architecture decision record for a gaming moderation assistant puts a policy stage between retrieval and generation, and counts abstention with a reason code as a successful outcome.
Reality
A dev.to post argues the "RAG or agents" question is malformed. The interesting part is the invoice: one axis out of five justifies the premium, and most buyers never use it.
Reality
A solo builder's writeup of an AI support SaaS puts the model behind a service boundary. Isolation, per-answer cost and freshness turn out to be architecture problems, not prompt problems.
Reality
A dev.to walkthrough wires chokidar, tree-sitter, Qdrant and an MCP server into a code search stack. Every part is off the shelf; the unresolved work is reindexing and retrieval quality.
Reality
A dev.to writeup drops LangChain, LlamaIndex, hosted vector stores and the cloud model, leaving six files, seven npm packages, and one cloud dependency still in place.
Reality
A dev.to build log for KnowBase AI publishes the load-bearing parts: retrieval scoped by workspaceId, chunks that keep a pointer home, and citations you can click. The retriever is the swappable half.
Reality