What it is
A framework for retrieval-augmented generation whose core value is ingesting data of any format and source: documents, databases, APIs, knowledge graphs, then organising it into queryable indices. Its abstraction layers are more complete than most peers.
Highlights
- Ingests documents, databases and APIs
- Offers layers from simple vector search to graph retrieval
- Many ready-made connectors included
- Query engine and chat engine designed separately
How to use
Install it, load a directory and build an index in a few lines, then swap retrieval strategies to tune results.
License
Released under MIT. Read the terms before commercial use or redistribution, especially if you plan to offer it as a service.