Synthetic Data Prototype: Built for AAYU 2026 Hackathon (Problem T01 • “The Two-Minute Doctor”). All clinical data is simulated.Privacy & Compliance →
DPDP Act 2023 CompliantABDM Milestone 2 Aligned

Privacy, Ethics & Clinical Governance Policy

Plain-language disclosure of how patient voice data, document OCR, and AI agent outputs are governed.

AAYU 2026 Hackathon Prototype Notice

Industra Medicose is demonstrated using 100% synthetic, anonymized clinical data created for Problem T01 “The Two-Minute Doctor”. No real patient Protected Health Information (PHI) is processed or retained without explicit consent.

1. What Data We Process

We collect audio microphone input during patient intake (transcribed via browser speech API), user-entered symptoms, and uploaded medical documents (prescriptions, lab tests). Speech audio is processed in memory and never permanently stored as raw audio files.

2. Doctor-in-the-Loop Mandate

Strict Clinical Posture: The AI agent system NEVER makes an unverified clinical diagnosis, never prescribes pharmaceuticals directly to patients, and never lowers emergency triage urgency. Every summary must be reviewed and digitally signed by a licensed physician before release.

3. Data Retention & Storage Security

Data is retained strictly for the duration of the clinical consultation lifecycle (30-day demo cycle). Documents uploaded to Supabase Storage are protected by private bucket Row Level Security (RLS) policies requiring authenticated token verification.

4. Consent & Revocation Rights

In accordance with Section 6 of the Digital Personal Data Protection Act (DPDP 2023), patients retain unconditional rights to withdraw consent at any time and demand instantaneous data erasure without consequence.

Exercise Right to Erasure (DPDP Act 2023)

Delete all synthetic consultation records, anonymize clinical intake transcripts, and purge document storage buckets associated with your current session.

© 2026 Industra Medicose. Built with empathy and precision for clinical practitioners.

AAYU 2026 Hackathon • Problem T01: The Two-Minute Doctor