// Personal project
I built a medical AI platform where users can chat about their symptoms, upload scans (CT, MRI, X-ray) for automatic analysis, and get medical reports explained back to them in plain language.
Given the stakes, I made sure a real professional stayed in the loop.
AI health tools that hand out recommendations directly to users, with no human checkpoint, are risky. I wanted the AI to help interpret and explain, but keep an actual medical professional in the loop before anything reaches the patient.
I built the scan-analysis and chat features to do the heavy lifting on interpretation, but deliberately routed every medicine recommendation through a doctor-verification step instead of letting the AI act unsupervised.