Inspiration

Presage prior health focus alongside my psychology class at college.

What it does

Takes the information provided by Pre-Sage and uses it while incorporating an A.I need to make up analyses and help the therapist deduce probabilities.It can be determined from a person's breathing patterns, speech, and facial expressions to figure out how the person reacts to certain topics/information. This helps therapists know what type of topics are sensitive for the person and what topics to ultimately target. It can help track trauma response over time as well, it is mainly for live action analysis of a persons facial expressions to find out how and which topics they respond to and how.

How we built it

We used the Pre sages api key alongside A. I need integration to have the information run through and give feedback.

Challenges we ran into

Accomplishments that we're proud of

Built a working prototype that integrates physiological context into therapy support. Successfully framed AI insights as assistive, not authoritative. Created a dashboard that therapists can understand quickly before or after a session. Avoided overclaiming while still demonstrating real clinical potential. Designed with privacy and consent as first-class features, not afterthoughts.

What we learned

Mental health AI must prioritize trust and interpretability over raw prediction power. Small, well-framed insights are more valuable than complex black-box scores. Therapists want context, not conclusions. Building responsibly in healthcare is as much a design problem as a technical one. Hackathons reward clarity of intent more than technical excess.

What's next for PrePysch

Add baseline calibration so each client’s responses are compared to themselves over time. Incorporate optional wearable trends (sleep consistency, resting HR, HRV) for between-session context. Build a session timeline view highlighting response peaks tied to topic shifts. Introduce therapist feedback tagging to refine insights and reduce false positives. Pilot the tool with a small group of therapists to evaluate real-world usefulness. Strengthen privacy controls and data minimization before any real-world deployment.

Built With

  • api
  • css3
  • presage
  • pythong
Share this project:

Updates