$ show lancedb/vectordb-recipes
This repository is a comprehensive collection of examples, tutorials, and applications demonstrating how to use LanceDB (a vector database) for various AI/ML applications. It showcases patterns for retrieval-augmented generation (RAG), multimodal search, chatbots, recommendation systems, and AI agents. The repository contains both Python and JavaScript/TypeScript implementations, with many examples featuring web interfaces using Streamlit, Gradio, or React. While it serves as a valuable resource for learning LanceDB integration patterns, the quality varies across examples, and some content is marked as archived.
The LLM identified 3 files worth highlighting for reuse:
This file implements a LanceDB adapter that makes it compatible with NVIDIA's RAG framework. It demonstrates the adapter pattern for integrating LanceDB with external systems, showing how to implement a vector database interface that can be swapped with other vector stores. The component is relatively self-contained and showcases non-obvious skill in framework integration.
This file contains a comprehensive set of metrics for evaluating RAG systems, including guard models for toxicity, bias, and prompt injection, string similarity metrics, and reference-based metrics like BERTScore, ROUGE, and BLEU. It provides a reusable toolkit for assessing RAG system quality that could be extracted and used in other projects with minimal modification.
This file implements an agent architecture using PydanticAI that incorporates LanceDB as a tool for recipe retrieval. It demonstrates tool-based reasoning patterns where agents can search a recipe database, conversational memory through vector storage of interactions, and multimodal query handling for text and image inputs. The agent pattern shows clean separation of concerns and reusable design patterns for AI agents.