MediGist : Secure AI powered medical record summarization for doctors and patients
Link to open source: https://github.com/Svizzcodes/TheSemicolonistsGDG.git
Link to Live Project: https://semicolonists-medigist.vercel.app/
MediGist is a privacy focused medical record summarization platform designed for doctors and patients. It converts long, unstructured medical documents into clear, role specific summaries using AI, while applying automatic redaction to protect sensitive personal information.
Doctors get clinically structured summaries that highlight conditions, medications, procedures, and key observations. Patients receive simplified, easy to understand explanations of the same records without medical jargon.
The system works entirely on user provided text or uploaded files, applies redaction before any AI processing, and maintains local patient history for continuity of care. MediGist emphasizes security, clarity, and practical usability over flashy features, making it suitable for real world clinical and educational use cases.
Key Features Include:
• Dual summaries
Doctor-focused clinical summary and patient-friendly simplified summary
• Privacy-safe processing
Client-side redaction, no permanent storage of raw records, stateless backend
• Secure access
Separate login for doctors and patients with role-based views
• Upload & history
Supports file upload and maintains patient history for continuity
• Explainability & trust
Summaries generated strictly from provided content with audit visibility
• Anti-misuse protection
Non-copiable summaries, right-click disabled, tab-blur protection
• Dynamic watermarking
Each summary is watermarked with user name, role, and access time
This build was uploaded as a hackathon project
