This repository contains updated code for WaitGPT: Monitoring and Steering Conversational LLM Agent in Data Analysis with On-the-Fly Code Visualization (ACM UIST 2024). The current version supports VLM-driven data analysis.

License: MIT

1. Copy the example config: cp token.yml.example token.yml
2. Edit token.yml with your OpenAI API key.

token.yml is gitignored and must not be committed.

The frontend is built with Node.js 22.11.0. It runs on localhost:5173 by default.

cd interface
npm install
npm run dev

The backend is built with Python 3.12.8. It runs on localhost:12050 by default.

cd server
pip install -r requirements.txt
python run.py

Start the server before the interface so the UI can load API configuration from the backend.

This project is intended for local development only. The server exposes GET /key, which returns secrets from token.yml to the browser. Do not deploy this stack to the public internet without removing that endpoint and using environment-based configuration.

Some code is adapted from open source projects. See NOTICE for details.

• OpenInterpreter: open-interpreter

@inproceedings {xie2024waitgpt,
    title = {WaitGPT: Monitoring and Steering Conversational LLM Agent in Data Analysis with On-the-Fly Code Visualization},
    author = {Liwenhan Xie and Chengbo Zheng and Haijun Xia and Huamin Qu and Chen Zhu-Tian},
    booktitle = {Proceedings of the 37th Annual ACM Symposium on User Interface Software and Technology},
    year = {2024},
    numpages = {14},
    articleno = {119},
    publisher = {Association for Computing Machinery},
    address = {New York, NY, USA},
    url = {https://doi.org/10.1145/3654777.3676374},
    doi = {10.1145/3654777.3676374},
    location = {Pittsburgh, PA, USA},
    series = {UIST '24}
}