About the Project
BART started from a simple idea: the art market is huge, but still surprisingly hard to navigate. Information is scattered, prices are unclear, and a lot of decisions rely on intuition rather than structured data. We wanted to see if we could make this world more understandable and easier to explore.
Through this project, we learned how to turn raw auction data into something meaningful. We worked on financial concepts like repeat-sales and hedonic regression, but also on building a full product around them, from backend APIs to a responsive frontend. It was a great opportunity to connect theory with real-world use cases.
We built BART as a complete system: a pipeline to process data, a backend to compute metrics and expose them, and a frontend designed to make exploration fast and intuitive. A lot of attention went into making the interface feel reactive and useful, not just technically correct.
The biggest challenges were dealing with messy and incomplete data, designing metrics that actually make sense for such an illiquid market, and making everything feel smooth despite the complexity behind the scenes.
In the end, BART is our attempt to make the art market a bit more transparent, and to show how data can support better understanding and decision-making in places where it’s usually missing.
Log in or sign up for Devpost to join the conversation.