PharmaFindr

AI-powered app that scans prescriptions, broadcasts in real time to nearby pharmacies, confirms stock, and lets patients reserve meds in a few taps—bridging healthcare gaps.

Comment

Inspiration

When my wife went into labor, we faced a terrifying moment — I had to leave her side in the middle of an emergency to search for a critical medicine. Pharmacy after pharmacy, I searched for over an hour, unsure who had stock, who was open, or where to go. That experience exposed a painful reality: in countries like the Philippines, there’s no centralized way to check medicine availability. People are left to walk store to store, even during emergencies. We built PharmaFindr to change that — so no one else has to go through what we did.

What it does

PharmaFindr is an AI-powered app that lets patients scan their prescriptions and instantly broadcast them to nearby pharmacies. Pharmacies confirm stock with a single tap, and patients can reserve medications in just seconds. The app provides real-time updates — from confirmation to ready-for-pickup — eliminating the guesswork and delays in getting essential medicine.

How we built it

AI-Powered Development: We used tools like Bolt.New and Claude 4 to accelerate development — from debugging and UI generation to backend logic and API calls.

Voice & Storytelling: Used ElevenLabs for natural voiceovers and Google Veo 2 to create a compelling, high-quality demo video.

Frontend: Built using React Native with Expo to support both patient and pharmacy interfaces in a single, efficient codebase.

Backend: Runs on AWS (Lambda, API Gateway, DynamoDB) for speed, scalability, and cost-effectiveness.

AI/ML: Integrated Google Gemini, a powerful multimodal AI model, to accurately read both printed and handwritten prescriptions.

Real-Time Features: Implemented real-time stock confirmation and reservation tracking between patients and pharmacies.

Challenges we ran into

Limitations with Bolt.New on large codebases Our combined codebase for both patient and pharmacy apps began to slow down and crash in StackBlitz. We solved this by linking to GitHub and working offline via Claude 4. We also separated the landing page (built with Bolt.New) for better performance and maintainability.

Accurately reading handwritten prescriptions We tested multiple AI/OCR models and found Google Gemini provided the most consistent and accurate results across a wide range of handwriting styles.

Real-time sync between patient and pharmacy apps We had no prior experience building real-time systems. With help from AI tools like Bolt.New and Claude 4, we implemented a functional live-update system between both apps for stock confirmation and reservation tracking.

Accomplishments that we're proud of

Built a fully working prototype in just weeks.

Successfully scanned and matched dozens of real prescriptions.

Designed a smooth, intuitive UX for both patients and pharmacies.

Deployed our landing page directly from Bolt.New to Netlify.

What we learned

Building with AI tools: Created a landing page and app using Bolt.New and Claude 4, speeding up design and dev time.

Easy deployment: Launched our landing page via Netlify directly from Bolt.New with minimal setup.

AI integration: Learned to integrate Google Gemini into our app for highly accurate OCR, even with difficult handwriting.

Real-time systems: Gained experience building real-time communication between two app environments (patient and pharmacy), a problem we had never tackled before.

Validating the Problem and Market Opportunity: As of 2025, the world population is approximately 8.1 billion (United Nations, 2024). According to the International Pharmaceutical Federation (FIP), there are an average of 2.78 community pharmacies per 10,000 people. Using this metric, there are an estimated 2.25 million community pharmacies globally (8.1B ÷ 10,000 × 2.78 = ~2,251,800 pharmacies). The global retail pharmacy market is valued at over $1.3 trillion USD, yet in many low- and middle-income countries, pharmacy access remains fragmented, offline, and inefficient—especially in moments of urgent need. In underserved regions like Southeast Asia, Africa, and Latin America, pharmacy infrastructure exists, but systems are not digitized, and real-time stock visibility is rare. This represents a $100–300 billion addressable opportunity for real-time, AI-powered solutions like PharmaFindr.

What's next for PharmaFindr

  1. Start working on compliance — Address healthcare regulations and data privacy laws to ensure trust and legal readiness.

  2. Partner with a barangay in Cebu, Philippines — Begin a real-world pilot with local government support to validate our solution in community healthcare settings.

  3. Gather data and feedback — Observe how users interact with the platform and use that feedback to improve UX, performance, and accuracy.

  4. Implement new features — Based on pilot learnings, we plan to add features like delivery tracking, offline access, and smarter pharmacy tools.

  5. Build monetization features — Introduce subscription tools for pharmacies — including analytics, inventory optimization, and premium search visibility — to ensure financial sustainability while keeping the service 100% free for patients.

  6. Scaled city rollout in 2026 — Launch PharmaFindr across major urban centers, starting with Cebu, Philippines. This marks our first full-scale deployment based on real-world pilot insights.

Built With

Share this project:

Updates