The-Vibe-Company/quivr
Opiniated RAG for integrating GenAI in your apps 🧠 Focus on your product rather than the RAG. Easy integration in existing products with customisation! Any LLM: GPT4, Groq, Llama. Any Vectorstore: PGVector, Faiss. Any Files. Anyway you want.
What it solves
Quivr provides a streamlined way to build a "second brain"—a personal AI assistant that can answer questions based on your own private documents. It removes the complexity of setting up Retrieval-Augmented Generation (RAG) from scratch, allowing developers to integrate a knowledge base into their applications quickly.
How it works
Quivr uses a RAG pipeline that can be configured via YAML files to define the workflow (e.g., history filtering, query rewriting, and retrieval). It supports a wide variety of LLMs (OpenAI, Anthropic, Mistral, Gemma) and local models via Ollama. Users can ingest various file types (PDF, TXT, Markdown) and customize the retrieval strategy, including adding internet search and external tools.
Who it’s for
Developers who want to add a document-based AI assistant to their project without managing the intricacies of the RAG pipeline.
Highlights
- Opinionated RAG: A fast, efficient, pre-built RAG implementation designed for rapid deployment.
- Flexible LLM Support: Compatible with multiple major AI providers and local models.
- Broad File Compatibility: Supports PDF, TXT, and Markdown, with the ability to add custom parsers.
- Configurable Workflows: Allows users to define the RAG process through YAML configuration files.
- Megaparse Integration: Works with Megaparse for enhanced file ingestion.
Related
- Project
- Project
- Project
- Project