Inspiration
The idea came after watching a talk at DataDrivenNYC, where ASH presented how AI is being used in therapy. It made me wonder: What if AI could help without replacing the human professional?
TriMind grew from that thought — a supervised, collaborative model where AI assists, the doctor reviews, and the patient stays supported.
React, Typescript, bolt.new, Supabase
Create a fast prototype that explores how AI can help the therapy process in a supervised, doctor-friendly way. The goal was not to finish a full product, but to validate the model, the tone, and the workflow.
The idea came after watching ASH present at DataDrivenNYC. Instead of fully autonomous chat therapy, I wanted a model where AI is inside the care flow, drafting insights, preparing notes, guiding check-ins, and the doctor stays in control.
The inspiration also came from common therapist pain points: admin overload, note-taking, intake fatigue, and patient follow-up. TriMind tests how AI can help with that while keeping the clinical voice safe and supervised.
The biggest challenge is doctor adoption. Pure AI chatbots are easy to use but often unsafe; fully manual workflows are trusted but slow. The balance is showing clinicians that AI inside the process works better than AI outside it.
TriMind needed a flow that reduces work, never increases it, and keeps doctors fully in control.
The solution was to design a supervised model:
AI gathers intake, drafts insights, and prepares summaries.
Any sensitive guidance goes into a review queue.
The doctor can approve, edit, or decline replies before they reach the patient.
Crisis keywords automatically route to human-only responses.
This keeps trust high, reduces admin work, and allows AI to help where it’s safe.
