Publish-Ready Product Demo_ | api documentation video made with ngram

About the video

This api documentation video was made with ngram to walk through ll-rag, a FastAPI retrieval service backed by Postgres and pgvector for building retrieval-augmented generation pipelines. The video shows spinning up the container, loading embeddings with the all-MiniLM-L6-v2 model, and calling the retrieve endpoint to return ranked chunks with similarity scores in milliseconds. It also shows running an evaluation script across 644 queries to print recall metrics instantly, demonstrating that the pipeline performs as expected. The video closes by pointing viewers to the project on GitHub so they can clone it and ship their own RAG API. The scenes move in a clear technical sequence from setup, to querying, to evaluation, to open-source access, giving developers a concrete look at the service's performance and workflow before they decide to integrate it into their own stack.