Quicknotes

This is a work-in-progress chat assistant for (primarily) research papers.

I built this using:

• Fastapi (Python)
• Next/React (Typescript)
• PostgreSQL
• Docker

Along with Llama-Index for RAG.

Getting Started

Requirements

• Docker >= 25.0.3
• UV, which you can install here
• npm >= 11.5.1

Running

1. Clone this repository:
2. Generate API keys for:
   • OpenAI
   • Mistral
   • Gemini
3. Make a copy of .env.template, rename it to .env, then fill in each environment variable.
4. In ./backend/, run uv sync to install the required python packages
5. In ./frontend/, run npm ci to install the required TS packages.

To run Dev mode, run the following in the project root: bash start_dev.sh

This will create a tmux window that has panes with:

• postgres docker container
• Next dev server
• Fastapi dev server

Usage

Once you start the dev server, open: http://localhost:3000/auth/create to create an account
or
http://localhost:3000/auth/login to login.

These accounts will be hosted on your machine, as I haven't gotten a chance to set up hosting on my VPS.