Built a full-stack AI healthcare assistant designed to help patients describe symptoms, receive general orthopedic guidance, and book appointments through a natural conversational interface. The goal was to reduce friction in the patient intake process and give clinics a scalable, intelligent front-end layer.

Orthopedic clinics deal with a high volume of repetitive patient inquiries, from symptom triage to appointment scheduling. Most existing solutions are either static FAQ pages or clunky form-based intake flows. The client needed something that felt conversational, was medically responsible, and could scale across multiple channels.

Conversational AI interface using a Next.js frontend with streaming responses for a real-time chat feel

FastAPI backend handling all AI inference, session management, and appointment scheduling logic

RAG-ready architecture so the assistant can be grounded in clinic-specific knowledge bases (treatment protocols, FAQs, doctor bios)

Structured patient intake that collects symptoms, duration, severity, and relevant history before routing to a human or booking slot

Emergency symptom detection that flags high-risk inputs and redirects patients to urgent care pathways

Appointment scheduling integration connected to clinic calendar systems

Frontend: Next.js, React, Tailwind CSS

Backend: FastAPI, Python

AI Layer: LLM integration with streaming, RAG pipeline architecture

Infrastructure: Scalable API design ready for WhatsApp and clinic CRM extensions

Reduced repetitive intake workload for clinic staff through automated first-contact handling

Emergency detection layer adds a meaningful safety net for high-risk patient queries

Architecture is modular and extensible — the same core can be deployed to WhatsApp, web widgets, or integrated into existing clinic CRMs

Designed with healthcare safety guardrails throughout, not bolted on as an afterthought