What is retrieval augmented generation (RAG)?
Retrieval augmented generation means looking the answer up before writing it. The system searches your own documents for the passages relevant to a question, hands those passages to the model, and asks it to answer from them. The model supplies the language. Your files supply the facts, and the answer can cite which ones.
Search your documents first, then let the model answer from them.
Why it matters to you
The reason to care is citation. A model answering from memory cannot tell you where an answer came from, so you cannot check it. A retrieval system can point at the paragraph, which turns an answer you have to trust into an answer you can verify in four seconds.
The quality of a retrieval system is almost entirely in the retrieval, not the model. If the search returns the wrong three passages, a better model writes a more convincing wrong answer. This is why the boring parts, chunking, metadata, permissions and freshness, decide whether the thing is useful.
Permissions are the part that gets skipped. A retrieval system that ignores who is asking will happily quote the salary review to the intern.
Where this turns up in something OCTYN runs
Each line carries the date it was recorded and where it came from.
- 2026-09-14
Company Brain cites the source of every answer it gives, and runs on Alchemyst's verifiable context engine
src/lib/services.ts
If a proposal used this word and did not explain it, ask.
Book a call →