Cognita
Modular open-source RAG framework for building production document retrieval applications.
About
Built by TrueFoundry, Cognita organizes retrieval-augmented generation code for production use rather than notebook experiments. It splits a RAG system into swappable components: data loaders, customizable parsers for many document types including audio and video, embedders, a metadata store, vector databases such as Qdrant and SingleStore, an LLM gateway, and API servers, with LangChain and LlamaIndex used under the hood. Incremental indexing avoids reprocessing documents that were already ingested, and a bundled web frontend lets non-technical users configure retrieval options like query decomposition and reranking, then chat with their documents without touching code. Everything is API driven and deploys locally with Docker Compose or on TrueFoundry's managed platform. The framework is open source, though the repository notes it is no longer actively maintained, so engineering teams now tend to treat it as a reference architecture for structuring their own production RAG stacks.
Reviews (0)
Leave a Review
No reviews yet. Be the first to review!
Details
- Category
- RAG & Document Retrieval
- Price
- Free
- Platform
- Local/Desktop
- Difficulty
- Intermediate (3/5)
- License
- Apache-2.0
- Added
- May 7, 2026
Related Tools
All-in-one desktop and Docker app for private LLM chat with your documents.
Open-source data extraction and indexing engine for RAG applications.
Open-source embedding database for AI applications
Web scraping API that turns websites into clean LLM-ready markdown.
Local RAG capabilities in GPT4All for chatting with documents privately.
Cloud-native vector database for scalable similarity search