Inspiration

Trading is a game of nerves. Every day a retail trader has to make thousand of decisions: when to buy, when to sell, when to hold. Despite us being in the Golden Age of data, with more market alpha than ever before, many struggle to stay profitable. It's not about a lack of information... it's about the human psychology.

National Bank provided a challenge for QHacks 2026 to develop a solution to trader biases. This is my submission.

What it does

It uses ML to present the user with a "bias" profile and action points. It meets the user where they're at: not just a dashboard, but it's available as an extension to trading view, as well as a morning brief email.

How we built it

Model trained with scipy + xgboost. Website built with Next.js + React + TailwindCSS. Email built with Resend.

Built With

Share this project:

Updates