Team Idea
**Autonomous Health AI** (about 980 characters)
Autonomous Health AI is a multimodal agent that sees, reasons, and acts on patient data. Clinicians and patients juggle scattered scans, lab PDFs, and symptom notes, and urgent findings get buried. Our agent reads medical images and scanned documents with a vision-language model (Qwen-VL), compares results against reference ranges and patient history stored in a vector memory, assigns a triage level (routine, follow-up, urgent), and takes action: drafting clinician summaries, scheduling follow-ups, or raising alerts via MCP tools. Every output includes confidence scores, cited evidence from the source, and human-in-the-loop review for urgent cases. Built with PyTorch, vLLM, LangChain, and Hugging Face Spaces, running on AMD Developer Cloud (MI300X) with ROCm, whose large memory handles high-resolution imagery and multi-document context at high throughput. It speeds up review and helps catch urgent findings earlier.
Team Members
Unknown Hacker
shimulrajdas00121428
Submission
Team Leader hasn't made a submission yet