Skip to content

Namerlight/14-AutoMate-HoyaHacks

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

3 Commits
 
 
 
 
 
 
 
 

Repository files navigation

14 - Auto-Mate

HoyaHacks 2024 - Shadab Choudhury & Naren Sivakumar

This repository contains the two separate codebases for the Android App Front-End of Auto-Mate and the Python FastAPI Backend.

Large sections of this code are based on the open-source release of AppAgent developed by Tencent [1]. We modify the original codebase to make it more robust and reliable, then build FastAPI API endpoints to connect to it. We also added additional verification steps such as using a GPT-4 call as a 'fuzzy-matcher' in case an user forgets to mention which app to use originally.

To run this section of the code, you will need a working OpenAI API key with access to the GPT-4o model. Each step costs approximately $0.03 and each task may take anywhere from 2-3 to 20+ steps.

Also this code doesn't actually work.

The android component is built using Expo and React Native. It does not communicate with the python backend can issue commands to a connected android phone with USB debugging enabled or to an emulated phone.

[1] Zhang C, Yang Z, Liu J, Han Y, Chen X, Huang Z, Fu B, Yu G. Appagent: Multimodal agents as smartphone users. arXiv preprint arXiv:2312.13771. 2023 Dec 21.

About

An Agentic AI System that uses an Vision-Language LLM to control a mobile device to carry out broad tasks given by an user.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors