MediBuddyAI — Health-Tech Agent
AI-first scheduling, follow-ups, and care insights system using RAG to assist doctors and patients. Built with Groq-deployed AI models and LiveKit-based Voice AI agent.
The Project
MediBuddyAI was my winning submission for the AAIMS25 AI Code Fest — a healthcare reimagining hackathon. The system addresses a core problem in under-resourced healthcare systems: administrative overload on doctors and fragmented patient follow-up.
The AI agent handles appointment scheduling, sends follow-up reminders, answers patient FAQs, and surfaces relevant medical history context to clinicians — all through a conversational voice interface built on LiveKit and a RAG-powered knowledge base.
Technical Highlights
RAG pipeline over medical FAQ and appointment data — enabling instant, context-aware responses to patient queries without manual lookup.
Groq-deployed LLMs for low-latency voice responses in clinical settings.
LiveKit voice agent with custom wake-word and turn-taking logic for natural doctor-patient conversations.
FHIR-compatible patient record summaries surfaced to clinicians at point of care.
Want to build a similar health-tech AI system?