Inspiration
We realized that AI models are often biased in invisible ways, and more and more stories are emerging about the real impact these biases have on people's lives. Determined to tackle this problem, we sought to create a technical solution that ML engineers could adopt while spreading awareness about this issue.
What it does
Glassbox is a built-in framework that provide real-time suggestions during model training. Our app focuses on an intuitive workflow that lets engineers accept fixes and suggestions right inside their workspace.
How we built it
Running on hope and free hackathon food, we used Python for the backend and Typescript for the frontend.
Challenges we ran into
It was a challenge to build an application that works inside an IDE, and the backend logic took many brainstorming sessions to fully develop.
Accomplishments that we're proud of
We are proud to have built a tool that pushes the boundaries of ai workflows aimed at solving a fast-growing problem in 2026.
What we learned
We learned that people learn and adapt the fastest when there are prizes on the line and a looming deadline.
What's next for Glassbox
We would like to implement faster processing logic + more polished UX.
Log in or sign up for Devpost to join the conversation.