An AI-powered solution to automatically perform actions on the computer based on the user query.

A project by Mohammad Tomaraei and Maqbool ur Rahim Khan.
Very simple! It has a cross-platform user client that talks to the API on a server with access to a visual large language model (VLLM).


Our aim has been to Dockerize every component of this project, but this is a work in progress due to compatibility issues with different platforms. On Windows and Mac, we recommend installing dependencies and run the server and client directly from your machine.
The following are the development dependencies, but once compiled, the client is a single binary for Windows, Linux, and Mac:
pyproject.toml files for specific version requirements)wails doctor to diagnose issuesStart by cloning the project:
git clone https://github.com/Stettiner-Valley/LetMeHelp.git
The test server is a simple Python script to create a WebSocket server and test the capabilities of the user client, such as receiving the user query and followup text messages, executing commands, and retrieving the data returned by commands.
To run the test server, open a terminal, navigate to server/src/server, run poetry install, and finally poetry run python -m server. The WebSocket server will be listening on 0.0.0.0:8765. Take a look at the comments in __main__.py and add your own commands to test bi-directional communication with the client.
You can also try running the LLM-powered server, but this is a work in progress. See the baserat directory. To run it, either use make run or poetry run python baserat/main.py.
To run the user client, open a terminal, navigate to user-client/src/letmehelp, and run wails dev. The client takes a few seconds to build, and will automatically reload when you make a change. See Wails Docs for more info.
user-client/src/letmehelp/app.go)If you'd like to contribute to our project, feel free to open a pull request :)
44 commits
12 commits
Jupyter Notebook
79.1%
Python
7.0%
Vue
3.5%
NSIS
3.1%
Go
2.0%
JavaScript
1.6%
Jinja
1.5%
An AI-powered solution to automatically perform actions on the computer based on the user query.

A project by Mohammad Tomaraei and Maqbool ur Rahim Khan.
Very simple! It has a cross-platform user client that talks to the API on a server with access to a visual large language model (VLLM).


Our aim has been to Dockerize every component of this project, but this is a work in progress due to compatibility issues with different platforms. On Windows and Mac, we recommend installing dependencies and run the server and client directly from your machine.
The following are the development dependencies, but once compiled, the client is a single binary for Windows, Linux, and Mac:
pyproject.toml files for specific version requirements)wails doctor to diagnose issuesStart by cloning the project:
git clone https://github.com/Stettiner-Valley/LetMeHelp.git
The test server is a simple Python script to create a WebSocket server and test the capabilities of the user client, such as receiving the user query and followup text messages, executing commands, and retrieving the data returned by commands.
To run the test server, open a terminal, navigate to server/src/server, run poetry install, and finally poetry run python -m server. The WebSocket server will be listening on 0.0.0.0:8765. Take a look at the comments in __main__.py and add your own commands to test bi-directional communication with the client.
You can also try running the LLM-powered server, but this is a work in progress. See the baserat directory. To run it, either use make run or poetry run python baserat/main.py.
To run the user client, open a terminal, navigate to user-client/src/letmehelp, and run wails dev. The client takes a few seconds to build, and will automatically reload when you make a change. See Wails Docs for more info.
user-client/src/letmehelp/app.go)If you'd like to contribute to our project, feel free to open a pull request :)
44 commits
12 commits
Jupyter Notebook
79.1%
Python
7.0%
Vue
3.5%
NSIS
3.1%
Go
2.0%
JavaScript
1.6%
Jinja
1.5%