# Keskin

- **Event:** [Built with Opus 4.7: a Claude Code hackathon](https://cerebralvalley.ai/e/built-with-4-7-hackathon)
- **When:** Apr 21 at 12:00 PM – Apr 27 at 2:00 AM (EDT)
- **Where:** Online
- **Placement:** 1st Place
- **Team:** [Bedirhan Keskin](https://cerebralvalley.ai/u/bedirhan)
- **GitHub:** https://github.com/bedriyan/medkit-app
- **Demo video:** https://youtu.be/6bN6hnx-A2A
- **Gallery:** https://cerebralvalley.ai/e/built-with-4-7-hackathon/hackathon/gallery
- **Page:** https://cerebralvalley.ai/e/built-with-4-7-hackathon/hackathon/gallery/28

Medkit is a virtual clinic that lets you make every mistake before the real patients come. It's a voice-first AI simulator for medical students and newly graduated doctors.

You manage scientifically generated AI patients in real-time voice conversation — take the history, order labs, read imaging, diagnose, and prescribe. The patients are real agents that simulate real people, with different personalities. Some are anxious, some are evasive, some downplay their symptoms, some ramble. You have to manage the communication as much as the medical reasoning — exactly like a real consultation.

After every session, Medkit grades your communication, history-taking, and clinical reasoning with scientific citations from the latest published guidelines (NICE, ESC, AHA, GINA, GOLD). The attending physician powered by Claude Opus 4.7 cannot fabricate guidance — its citations come from a curated registry, so the marking sheet quotes the same evidence a real examiner would cite. Trainees can run a dozen cases an hour and get a cited, structured debrief on every one. That feedback loop doesn't exist anywhere else.

---

Markdown version of https://cerebralvalley.ai/e/built-with-4-7-hackathon/hackathon/gallery/28. Site index for agents: https://cerebralvalley.ai/llms.txt · full text: https://cerebralvalley.ai/llms-full.txt