Agentset

Say you need a chat assistant that answers questions from thousands of contracts or medical documents and cites the page each answer came from.

Agentset is an open-source RAG infrastructure tool that lets developers build that kind of AI chat and search application without writing the retrieval pipeline themselves.

It covers document extraction, chunking, hybrid search and agentic reasoning, and the stated aim is to keep answers reliable with large data sets and heavy user traffic.

Agentset is model-agnostic, so you pick your own LLM, vector database and embedding model.

It reads more than 22 file formats, including multimodal content such as images and tables, and returns cited answers out of the box.

The examples it points to are medical AI, legal tech and enterprise search.

Pricing is enterprise, so expect to talk to the team rather than sign up on a fixed plan.

Because the code is open source, you can inspect how retrieval works before committing.