Show HN: I built an offline open-source RAG system DataBridge

[Adityav369]

https://github.com/databridge-org/databridge-core

Hey HN!

Over the past few weeks, I’ve been working on DataBridge, an open-source solution for data ingestion and querying across text, PDFs, images, and videos.

In our latest update, we’ve added a fully local deployment option:

- No internet required – Runs entirely offline. - Customizable Models – Supports any LLM and embedding model via Ollama (with options for any other private providers) - Extensibility – You can plug in your own models or tools easily.

This local-first approach ensures better privacy, security, and flexibility, especially for teams dealing with sensitive data.

You can check it out here:

Docs: https://databridge.gitbook.io/databridge-docs Code: https://github.com/databridge-org/databridge-core

If you can try it out, would really really appreciate it :)). Any and every feedback welcome.

[Haeuserschlucht]

How does it compare to notebooklm in terms of quality and length of output given llama3.2-8B? You need to give more context, currently it seems like this is yet another reinvention of the wheel.

[Adityav369 → Haeuserschlucht]

Good question! You can use any open source or private model you want with this (just by changing one line in `databridge.toml`), whereas notebook lm is limited to gemini. While our UI component allows it to be used as an open source notebook lm alternative, it is developed for developers building AI apps and doing context management. We are working on adding techniques like CAG (cache augmented generation) and graphRAG for better and more flexible querying, and have custom embedding pipelines for different data types.