IT-Mobile | Devpost

logo

Inspiration

We noticed that many users on T-Mobile’s forums and review pages frequently complain about poor Wi-Fi connectivity or router issues, often waiting hours to reach customer service for basic troubleshooting. Many users also don’t fully understand how to make the most of their router’s capabilities. We wanted to create a solution that can instantly identify and fix common Wi-Fi problems in an attempt to be as helpful and accessible as a “tech agent” that’s faster than calling support.

What it does

Run wifi speed checks.
Identify connectivity bottlenecks and suggest optimizations.
Easy accessibility, speech to text, and text to speech.
AI Powered.

How we built it

Using Raspberry Pi 5, T-Mobile 5G network. Backend using Python, Flask. Frontend using HTML, CSS, JS, Streamlit. Mongo DB. Statsig for model optimization. Also used Git, Command-Line, Linux. A RAG with Knowledge Base

AI/ML Stack

Transformers, PyTorch, Qwen, TinyLLama, Scikit-learn, NumPy

Challenges we ran into

Choosing a fast enough model, real-time wifi speed updates, speech recognition.

Accomplishments that we're proud of

We're extremely proud of using the Rasberry Pi and deploying our code on there so it can perform the troubleshooting tasks. We were proud of integrating CLI so the bot can fix an issue with the respective CLI commands because that plays a key part in this project. We also decided to make our project accessible to many users by implementing a Text-to-Speech API and a Speech-to-Text API.

What we learned

Each of us learned a high level school through this project. or example, working with Raspberry Pi, integrating CLI, working with AI powered model on an SLM.