Hey HN! We've just launched Quilt, a robust RAG (Retrieval-Augmented Generation) UI that revolutionizes how you interact with your documents.
Key features: - Multi-user setup with private/public document collections - Advanced hybrid RAG pipeline combining full-text & vector search - Smart citations with in-browser PDF preview and highlights - Fully customizable settings and prompts through the UI
Making an account is free, no need to even use a strong password: this is only to ensure your documents are separate from the rest.
We're keen to hear your thoughts and feedback. What features would you like to see next?
Can you explain how is this different from the other offerings. I believe there are quite a few RAG solutions out there, but its not all clear from your description how you guys stand out.
this is fork of https://github.com/Cinnamon/kotaemon Please attribute accordingly
I can see, use, and download files that have been uploaded by other users. I recommend to document that any file that is uploaded to the system can be used and downloaded by others.
This is missing an explanation for the “powerful” part.
No need to make a strong password? I’d think it be a good idea for private documents no?
Also the attribution issue, really means you have to explain what’s going on here.
Are there plans to provide a version that can be run locally? I would not be comfortable putting this on a remote server.
Are you going to be open sourcing this or do you have a pricing plan for later versions?
If this is to be non local then what security will be added to this?
Hi, I am not affiliated with the OP, but I would like to understand what your concerns are in the context of data security and privacy. Can you please also explain what is expected of AI companies in this aspect, especially for B2B applications?
Are you being serious?
Is this sota? How does rag quality compare to that in, say, lmstudio?
Also is there a "rag quality benchmark" yet
I'm curious as to how you achieve the document highlighting? Are you making a guess based on extracted text or do you use a vllm?
The about section says that it was built for developers as well as users. What part of it is useful for devs?
Is this a fork of kotaemon[0]? Kotaemon is apache licensed so would hate to see another PearAI fiasco.
This indeed seems like a fork of kotaemon.
I noticed that the server was experiencing some issues, it is all worked out now! Please comment here with any issues you encounter and I'll try to address them as soon as I can.
Are you able to integrate analysis of figures such as charts? I'd really like to be able to extract data such as differences between data points (absolute and relative, with correct units), and analyze trends.