Your Medical Companion

Inspiration

This project was inspired by the communication gap that often exists between patients and caregivers after medical visits. Patients frequently struggle to recall everything a doctor says, while caregivers may be left in the dark. We wanted to build an AI-powered solution that captures, organizes, and shares this critical information clearly and securely.

What it does

Our Medical Companion platform allows patients to record or upload doctor conversations.
AI then transcribes these recordings, extracts key details, and generates structured medical reports.

Caregivers can:

• Securely access transcripts and reports
• Schedule AI-powered phone calls with patients

During these calls, the AI agent:

• Conducts the conversation with the patient
  
• Gathers updates about their health or care needs
  
• Produces a call summary for the caregiver, enabling proactive care

How we built it

• Supabase for authentication, database management, and secure storage of audio files and transcripts
  
• Whisper (ASR model) for multilingual audio-to-text transcription
  
• AI summarization pipeline to generate patient-friendly and caregiver-focused reports
  
• Supabase Edge Functions for scheduling and managing AI-powered phone calls
  
• FFmpeg for normalizing audio formats like WAV, MP3, and M4A

Challenges we ran into

• Handling multiple audio formats and ensuring consistent quality for transcription
  
• Ensuring accurate transcription in medical contexts, especially with technical terminology
  
• Designing simple interfaces for patients while providing detailed dashboards for caregivers
  
• Implementing time-zone aware scheduling for AI-powered calls
  
• Balancing data security (HIPAA/GDPR compliance) with ease of use

Accomplishments that we’re proud of

• Building a real-time system that captures and clarifies medical conversations
  
• Enabling caregivers to stay connected through AI-powered phone check-ins
  
• Creating a dual-user experience, where patients and caregivers see personalized, role-specific dashboards
  
• Establishing a strong foundation for a compliance-ready healthcare platform

What we learned

• The importance of user accessibility in healthcare design
  
• How to integrate ASR (Automatic Speech Recognition) models with structured data pipelines
  
• The value of Supabase Row-Level Security (RLS) for protecting sensitive health data
  
• How proactive care tools, like scheduled calls, increase caregiver confidence and patient safety

What's next for Medical Companion

• Compliance & Security: Achieving HIPAA/GDPR readiness
  
• Smarter Reports: AI that not only summarizes but also identifies risks, reminders, and trends over time
  
• Care Coordination Tools: Shared calendars, medication reminders, and symptom trackers
  
• Multilingual Support Expansion: Serving global patient–caregiver communities
  
• Integration with Healthcare Systems (EHR/EMR): Making the platform interoperable with existing clinical workflows

Built With

• claudecode
• next.jsopenaisdk
• supabase
• vapi.ai