ClinicCall AI

Dashboard
Dashboard

Inspiration

I wanted to solve a real problem that clinics face: staff spending valuable time making repetitive phone calls to remind patients about appointments.

I saw an opportunity to use AI-powered phone calls to automate this process while keeping the experience simple for clinic staff. That idea became ClinicCall.

What it does

ClinicCall is an AI-powered clinic communication system that helps automate patient appointment reminder calls.

A clinic can manage patient information, select a patient, and initiate an automated phone call. CALL-E handles the voice conversation, and ClinicCall records the outcome so the clinic can keep track of completed calls.

I also added phone-number masking in the interface to help protect patient contact information.

How I built it

I built the frontend using React and Vite and the backend using Python, FastAPI, and SQLModel.

I integrated the CALL-E SDK into the backend so that ClinicCall can create and wait for real phone calls at runtime.

The workflow is:

React frontend → FastAPI backend → CALL-E SDK → Phone conversation → Call result → Call history

I also added a demo mode so ClinicCall can be demonstrated without making a real phone call.

Challenges I ran into

One challenge was handling phone-number formatting and CALL-E's regional calling restrictions. I learned how to use E.164 formatting and successfully tested the real CALL-E integration using the available US testing route.

Another challenge was making sure patient phone numbers were not unnecessarily exposed in the interface. I solved this by masking phone numbers while keeping the information available internally for authorized call operations.

I also had to make sure the frontend, backend, database, and CALL-E integration worked together as one complete workflow.

Accomplishments that I'm proud of

I'm proud that I turned ClinicCall from an idea into a working application with a real CALL-E integration.

I successfully made a real CALL-E test call and received a completed structured result from the conversation.

I'm also proud that I built the project independently and worked through the technical challenges of connecting the frontend, backend, database, and voice AI system.

Most importantly, I'm proud that ClinicCall is focused on solving a practical problem rather than being just a technology demonstration.

What I learned

I learned how to integrate a voice AI SDK into a real application and connect it to a frontend and backend workflow.

I also learned more about API integration, phone-number formatting, error handling, call results, privacy, and testing real voice applications.

Most importantly, I learned that building a useful AI product is not only about making the technology work. It is about identifying a real problem and creating a simple solution that can provide value to real users.