Inspiration Medication safety is still largely “one‑size‑fits‑all,” even though genetics and real‑time vitals can drastically change how a drug affects someone. We were inspired by how many preventable adverse drug reactions happen each year and wanted to give patients and clinicians a personalized, real‑time safety check before the first dose.
What it does MedGuard is an iOS app that combines a user’s 23andMe genetic data, their current medication list, and live vitals to produce a personalized drug‑safety assessment. It flags dangerous drug‑drug and gene‑drug interactions, assigns a clear traffic‑light risk level (green/yellow/red), and generates an AI‑powered report with actionable recommendations. If risk is elevated, it also supports follow‑up vitals monitoring to detect changes.
How we built it We built MedGuard in Swift/SwiftUI. The app parses 23andMe raw DNA data to extract key CYP450 variants, manages a medication list, and calculates an activity score that blends genetics, demographics, and lifestyle factors. We integrated the Presage SDK to measure heart rate and breathing rate via the iPhone camera. For user‑friendly guidance, we call the Gemini API to generate a concise, clinician‑style safety report. All processing runs client‑side for the hackathon demo.
Challenges we ran into Mapping raw genetic variants to meaningful metabolizer risk without over‑simplifying. Integrating real‑time vitals capture with a smooth SwiftUI experience. Balancing clinical depth with a UI that stays understandable for non‑experts. Time constraints: we opted for a client‑only demo instead of a full backend. Accomplishments that we're proud of End‑to‑end flow from DNA upload → medication input → analysis → AI report. Real‑time, camera‑based vitals integrated into a medication safety workflow. Clear, actionable traffic‑light risk system that’s easy to communicate. A polished iOS demo that feels like a real product, not just a prototype. What we learned We learned how to translate complex pharmacogenomics into a user‑friendly experience, how to combine multiple risk signals into one actionable output, and how to integrate biosensor data into a health‑tech product under hackathon time pressure.
What's next for MedGuard Add a backend service for larger drug‑interaction databases and safer key management. HealthKit and wearable integration for continuous vitals trends. Clinical validation with pharmacists and physicians. HIPAA‑grade privacy controls and secure sharing with care teams. Push reminders for follow‑up monitoring when risk is elevated.
Log in or sign up for Devpost to join the conversation.