AyushmanAI : The Intelligent Medical Assistant
Proposal
Overview
We propose an AI-powered assistant to transform the healthcare patient journey in India. The process of booking appointments, sharing patient history, and handling follow-ups can be streamlined, reducing friction for both patients and doctors. Our solution is two-part: a pre-consultation voice agent for efficient patient intake, and a post-consultation assistant that logs and summarizes conversations, empowering both sides with instant, reliable information.
Pain Points
- Patients often forget critical details during visits.
- Doctors spend excessive time on administrative tasks, leading to longer sessions.
- Patients seeking answers online risk misinformation.
By leveraging Agentic AI, our system becomes a proactive partner, not just a chatbot, addressing these gaps in care delivery.
How We Plan to Solve This
- Automated Patient Intake: A voice agent handles appointment calls, collects symptoms and history, and produces a concise summary for the doctor.
- Real-time Documentation: Consultations are transcribed in real-time, generating structured reports that replace manual paperwork and create accurate digital records.
- Voice-Activated Knowledge Retrieval: A RAG system enables doctors to ask voice queries on a patient’s history, ensuring quick access to critical details.
- Reliable Post-Consultation Support: An AI assistant offers accurate, personalized answers to patient questions after the visit, reducing reliance on unverified sources and improving treatment adherence.
Impact
This solution demonstrates an end-to-end automation workflow, combining IBM’s Granite models, preprocessing toolkits, performance benchmarks, and the Agent Development Kit (ADK). Beyond efficiency, it brings meaningful impact in an India-centric context where healthcare systems often struggle with scale, accessibility, and trust. By addressing intake, documentation, and patient support holistically, our proposal showcases how Agentic AI can reduce burdens on doctors, empower patients with clarity, and improve healthcare outcomes.
Built With
- asr/tts
- cloud-object-storage
- docker
- fastapi
- fhir-data-schema
- ibm-agent-development-kit-(adk)
- ibm-granite-models
- javascript
- kubernetes
- postgresql
- python

Log in or sign up for Devpost to join the conversation.