Autorag

Inspiration

I wanted a convenient way to scrap useful info from the internet. I feel it hectic to keep on providing context to my local LLM system. Making this a seamless experience was my motive.

What it does

You can scrape info from the internet and have your own personal rag model built on the information you collected.

How we built it

Supabase- Database Extension- Node.js App - React/Typescript

Challenges we ran into

• Making an extension interact with the webapp was very challenging.
• The "keep" feature wouldn't work on pdfs either at first.
• It was a lot of effort to allow multiple people edit corpus in the SAME session started by another.

Accomplishments that we're proud of

• I can build a Rag system and interact with it on the cloud. I can sign in on any device and my corpus is still accessible. This is a useful tool.

What we learned

• First time making an extension
• How to built a RAG ecoystem
• How to get better with WebMCP

What's next for Autorag

Built With

• node.js
• pgvector
• react
• supabase
• typescript