README
¶
14 · Semantic RAG — real embedding model + LLM
The embedder-backed counterpart to the lexical examples. It wires a real embedding model into CortexDB so retrieval works on meaning, not keyword overlap, and uses a chat LLM to answer — both via an OpenAI-compatible endpoint (the defaults target DashScope / Qwen).
docs ──InsertText (embeds via text-embedding-v4)──▶ CortexDB vector index
query ──SearchText (embeds query, cosine search)──▶ top hit by MEANING
│
qwen3.7-plus answers from that context
Why it matters
The query "who deals with production outages after hours?" shares no keywords with the stored fact "When the site goes down at night, the engineer on rotation gets paged…" — lexical/FTS retrieval would miss it. With an embedder, vector search ranks it first:
=== semantic vector search (no keyword overlap): "who deals with production outages after hours?" ===
1. [doc-oncall score=0.545] When the site goes down at night, the engineer on rotation gets paged and must respond.
2. [doc-payments score=0.296] ...
=== qwen3.7-plus answer ===
The engineer on rotation gets paged and must respond when the site goes down at night.
Run it
Requires an embedding + chat endpoint (config via env or a .env loaded by godotenv):
OPENAI_API_KEY=sk-... # required
OPENAI_BASE_URL=https://dashscope.aliyuncs.com/compatible-mode/v1
OPENAI_MODEL=qwen3.7-plus # chat
EMBED_MODEL=text-embedding-v4 # embeddings
EMBED_DIM=1024
go run ./examples/14_semantic_rag
Any OpenAI-compatible endpoint works (OpenAI, DashScope, a local Ollama with an
embedding model, etc.) — embedder.go is a ~60-line cortexdb.Embedder
implementation you can copy.
How it plugs in
embedder.goimplementscortexdb.Embedder(Embed/EmbedBatch/Dim).cortexdb.Open(cfg, cortexdb.WithEmbedder(e))withcfg.Dimensions = 1024switches CortexDB from lexical to vector retrieval.db.InsertTextembeds + indexes;db.SearchTextembeds the query and runs cosine similarity over the HNSW index.
Documentation
¶
Overview ¶
Semantic RAG — CortexDB with a real embedding model (vector search), plus an LLM answer. This is the embedder-backed counterpart to the lexical examples: it wires a DashScope (Qwen) text-embedding-v4 embedder so retrieval works on MEANING, not keyword overlap, and uses qwen for the final answer.
Reads config from the environment (or a .env via godotenv):
OPENAI_API_KEY (required) OPENAI_BASE_URL default https://dashscope.aliyuncs.com/compatible-mode/v1 OPENAI_MODEL chat model, default qwen3.7-plus EMBED_MODEL embedding model, default text-embedding-v4 EMBED_DIM embedding dimension, default 1024 go run ./examples/14_semantic_rag
Source Files