Mediguard

One-minute check, because every life matters

Comment

Inspiration

Medication errors are a silent global threat — they harm millions every year, often due to simple oversights like drug interactions, allergies, or misprescribed doses. We were inspired by the need to protect lives through prevention, not just reaction. MediGuard was born out of a desire to make healthcare safer, especially in areas where pharmacists are overworked or patients lack access to proper counseling. Every life matters — and sometimes, a one-minute check can make all the difference.

We also realized that veterinary medicine faces similar challenges — many pet owners rely on over-the-counter medications without proper guidance. That’s why we're planning a dedicated version of MediGuard for vets, to protect the lives of animals as well.

What it does

MediGuard is a smart assistant designed to help pharmacists and patients avoid medication errors. Users input basic patient information and prescribed medications, and MediGuard evaluates potential risks like drug interactions, allergy triggers, or conflicts with health conditions. The assistant uses structured logic and AI-style prompts to simulate pharmacist-like safety checks, making the experience intuitive and responsive. It also stores anonymized case statistics, offers quick alerts, and supports sharing cases via smart links — all without requiring login or user accounts.

How we built it

We built MediGuard using Bolt.new to structure the front-end logic and flow. For the backend, we used Supabase to manage anonymous case storage and apply Row-Level Security (RLS) to ensure that each session’s data remains isolated and private — even without requiring user authentication. The AI assistant component is powered by conversational logic that mimics intelligent feedback based on user input — while not a dedicated machine learning model, it behaves like a lightweight AI assistant. To enable features like case sharing, we deployed the app on Netlify, bypassing Bolt’s file-handling limitations through GitHub integration.

Challenges we ran into

  • Bolt.new’s rebuild process removed important redirect files, breaking case sharing — we solved it by deploying via GitHub and Netlify instead.
  • Implementing Supabase RLS policies correctly took time and precision, especially when configuring them for anonymous sessions without user authentication.
  • Designing a smooth conversational AI that doesn't overwhelm the user required careful prompt design and UI structuring.
  • Limited Bolt tokens restricted us from adding more advanced features we originally planned.

Accomplishments that we're proud of

  • We successfully implemented full case sharing through unique links.
  • Built a functioning AI-style assistant that evaluates drug safety in real-time.
  • Designed a clean, user-friendly interface for pharmacists and non-experts alike.
  • Integrated a secure Supabase backend with proper privacy protections — all without needing login.
  • Created something with real-world application that could truly save lives.

What we learned

  • How to work around platform limitations creatively (like hosting issues on Bolt).
  • Real-world application of Supabase RLS policies with anonymous users.
  • Importance of user-centered design in healthcare apps.
  • How critical even the smallest features (like reset buttons or alerts) are to UX in a high-stakes environment.

What's next for MediGuard

  • Integrate a real-time drug database API to auto-update medication information.
  • Add voice input support for accessibility.
  • Expand to multi-language support, starting with Arabic and French.
  • Implement AI-powered suggestions for safer alternatives based on region and patient history.
  • Launch a dedicated MediGuard for Vets version to help pet owners and veterinary professionals avoid medication errors in animals.
  • Explore partnerships with pharmacies and healthcare providers to pilot the tool in real-world settings.

Built With

Share this project:

Updates