Can this function as a frontend for my custom built RAG running a chromadb collection inside docker? #1154
Description
Problem is most chatbot UIs force that the API request is styled like OpenAI is, which is way to much for me, and to be honest I really don't feel like rewriting that part from the cloned repo.
I just need something pretty that can preferably be ran in docker, ideally comes with its own docker-compose yaml which i will then connect with my RAG inside another container on the same network.
I see that most popular solutions did not implement a simple plug and play with your own vector db, and that is something that i find out far too late when searching through github issues when i already cloned the repos.
So i decided to just treat the possible UI as a glorified curl like request sender.
I know i can just run the projects and add the documents as I go, problem is we are making a knowledge based solution platform for our employees, which I got to great lengths to prepare an adequate prompt, convert the files to markdown with markitdown and chunk with langchain markdown text splitter, which also has a sweet spot to grab the specified top_k results for improved inference.
The thing works great, but I can't exactly ask non-tech people to query the vector store from my jupyter notebook :)
I am not that good with frontend, and barely dabbled in JavaScript, so I hoped there exists an alternative, one that is straight forward, and won't require me to go through a huge codebase which I would need to edit to fit my needs.
Thank you for reading.