Dr. Chopper's Clinic

Dr. Chopper is a health assistant that knows when advice isn’t enough and helps users find clinics, call for availability, and choose an appointment—bridging AI guidance and real care.

Comment

🩺 Dr Chopper's Clinic – From Health Advice to Real Appointments

Inspiration

People already turn to LLMs for health advice, symptom explanations, and reassurance. But when the conversation reaches the point where real medical care is needed, today’s assistants hit a wall. They can tell you to see a doctor — but they can’t actually help you do it.

Dr Chopper's Clinic bridges that gap.

We built a personal health assistant that not only understands when a user should book an appointment, but can also take action on their behalf — finding nearby clinics, calling them, checking availability, and helping finalize the booking.

What it does

Dr Chopper is a conversational health assistant with agency.

1. Health-aware conversation

  • Uses an LLM to support users with health-related questions.
  • Continuously analyzes the conversation to detect moments where self-care is no longer enough and a clinic visit is recommended.

2. Appointment intent detection

  • Instead of passively suggesting “see a doctor,” the system recognizes intent signals such as symptom severity, duration, and escalation.
  • Transparently asks the user for consent before proceeding.

3. Clinic discovery & calling

  • Finds nearby walk-in clinics based on the user’s location.
  • Initiates a phone call to clinics using telephony and text-to-speech APIs.
  • Asks about availability on the user’s behalf (hours, next open slots, walk-in rules).

4. Human-in-the-loop booking

  • Aggregates responses from clinics and presents them clearly in chat.
  • Lets the user choose their preferred option.
  • Confirms the booking flow without wasting clinic staff time.

5. User profile & memory (bonus)

  • Optional user accounts.
  • Securely stores preferences and basic health context to personalize future interactions.

6. Live voice interaction (extra bonus)

  • Supports voice-based conversations so users can talk naturally instead of typing — especially useful in stressful health situations.

How we built it

  • Gemini AI for medical conversation support and appointment-intent detection
  • LLM logic to safely decide when escalation is appropriate
  • VAPI API for outbound clinic calls
  • Google Cloud Text-to-Speech (TTS) for natural, polite automated calling
  • Google Maps API location services for clinic discovery
  • React frontend for chat and voice interaction
  • Cursor was used for initial project structure

Challenges we ran into

  • Designing a system that knows when not to act
  • Avoiding false urgency or unsafe medical recommendations
  • Keeping clinic interactions respectful, short, and efficient
  • Balancing automation with explicit user consent

What we learned

  • LLMs become far more useful when paired with real-world actions
  • Healthcare workflows require trust, transparency, and human control
  • Voice interfaces significantly improve accessibility in high-stress situations

What’s next

  • Integration with real scheduling systems (where permitted)
  • Smarter follow-ups and reminders
  • Multi-language support
  • Stronger privacy controls and auditability

Disclaimer

This project does not book real appointments and is built as a prototype only. All calls and clinic interactions are simulated or conducted with explicit awareness and respect for healthcare providers’ time and services.

Built With

Share this project:

Updates