Quivr is an open-source full-stack Retrieval-Augmented Generation (RAG) platform. Our mission is to build the best stack that empowers anyone to create powerful RAG applications effortlessly.
We are developing a suite of tools:
- Megaparse: An open-source document ingestion tool for efficient data parsing and preprocessing.
- Quivr: Our core RAG engine that facilitates seamless retrieval and generation of information.
- Le Juge: An evaluation framework to assess and improve the performance of RAG applications.
- 🗂️ Document Ingestion with Megaparse: Easily ingest and preprocess large volumes of documents for your RAG applications.
- 🔍 Robust Retrieval with Quivr: Implement state-of-the-art retrieval techniques to enhance your AI models.
- 📊 Performance Evaluation with Le Juge: Evaluate and benchmark your models to ensure optimal performance.
- 🛠️ Full-Stack Solution: End-to-end tools covering ingestion, retrieval, generation, and evaluation.
- 🌐 Open Source Collaboration: Join our community and contribute to cutting-edge RAG development.