Recurvai

2025

Fullstack Developer

An AI-powered telemedicine assistant that conducts patient anamnesis, checks symptoms, and provides health recommendations. The platform uses OpenAI's language models to create conversational medical interviews, helping doctors triage patients more efficiently.

Location

United States (Remote)

Duration

2025

Team Size

15 members

Visit Live Site

My Contributions

Built the AI chat interface for patient anamnesis with streaming responses
Implemented the symptom checker engine with structured data extraction from conversations
Created the authentication and user management system with Privy auth to support blockchain-based identity verification
Store user data and history chat in Supabase PostgreSQL database with real-time updates for doctor-patient interactions
Integration with conversational API based on AI models from OpenAI for generating anamnesis questions, symptom analysis, and health recommendations

Challenges

1Implementing streaming chat UI with proper error handling for long-running AI completions
2Implementing animations and UI feedback to make the AI assistant feel responsive and engaging during conversations

Technology Stack

Next.js

Full-stack framework with API routes for OpenAI proxy and App Router for the patient/doctor interfaces

Shadcn UI

Accessible component library for building the chat interface and dashboard components

Supabase

Backend-as-a-service providing auth, PostgreSQL database, and Realtime subscriptions

OpenAI

GPT models for conversational anamnesis, symptom analysis, and health recommendation generation

Privy Auth

Authentication and user management system with blockchain-based identity verification