MedKnight

Context

MedKnight was a CalHacks 11.0 project built around an augmented-reality medical assistant. The pitch was to make medical guidance easier to follow by putting the assistant in the same visual context as the user instead of keeping everything in a separate chat window.

This was a hackathon project, so the focus was a working demo path rather than a complete clinical product. The interesting part was deciding what belonged in the AR layer, what belonged in the AI layer, and how to keep the interaction simple enough to explain quickly.

What it did

• Used an AR interface as the main interaction surface
• Combined AI-assisted guidance with a medical-assistant workflow
• Kept the user flow narrow so the demo could be tested quickly
• Focused on clear guidance instead of adding extra features

Result

MedKnight won at CalHacks 11.0. The useful part of the project was not that it solved healthcare. It was a fast prototype that made the interaction model concrete: AR for context, AI for guidance, and a workflow simple enough to demo under hackathon constraints.

Built in October 2024 with a small team.