🌍 Inspiration
I love traveling and often work remotely in different countries. But when I got sick abroad, I realized how difficult it was to navigate the healthcare system.
Questions like:
What does my insurance cover? Where should I go? Do they speak English?
were overwhelming — especially when I wasn’t feeling well.
That experience inspired me to build HealthAidFinder: a smart assistant that helps travelers find the right care quickly, confidently, and safely — like having a local nurse in your pocket.
💡 What it does
HealthAidFinder is a travel-focused AI health assistant that helps users:
- 🩺 Triage symptoms (mild, moderate, severe)
- 🧾 Check insurance coverage for different care types
- 🏥 Find nearby clinics or pharmacies based on urgency and location
It’s designed to make healthcare access during travel less confusing, especially when users are far from home and unsure what to do.
🛠️ How I built it
I built the project in two main parts:
🔧 Frontend (Bolt.new)
The UI was built using Bolt.new. I prompted the platform to create:
- A structured flow with different pages (e.g., main, profile, results, chatbot)
- Upload and location input tools
- User-friendly visuals and responsive design
Although Bolt.new doesn’t allow custom backend API integration, it was an intuitive no-code builder that helped bring the concept to life visually.
🤖 Backend (Bolt.new + Python + OpenAI API)
- The AI Nurse agent was developed in Flask, using OpenAI’s API.
- It includes tools for symptom triage, coverage checks, map search, and claim support.
- I deployed it publicly using Render so it could be accessed via URL.
- To work around the integration limitation, I embedded the chatbot as an iframe inside the Bolt.new frontend.
I used ClickUp for project planning, and ChatGPT to help brainstorm and break down development steps.
🚧 Challenges I ran into
No backend integration in Bolt.new
I spent a lot of time trying to connect frontend and backend, only to realize that Bolt.new doesn’t support API calls or code injection.Zero technical experience
I had no prior experience building apps, so I had to learn everything from scratch — including deployment and bug fixing.Time constraints
As a solo builder, I had to juggle learning, designing, building, and testing — all within the hackathon deadline.
🏆 Accomplishments that I am proud of
- Built a fully working AI-powered backend chatbot that can handle triage, insurance checks, and facility recommendations
- Designed and launched a frontend interface using Bolt.new with thoughtful UI/UX
- Found a creative workaround (iframe) to bridge the backend and frontend without full integration
- Learned and deployed everything from scratch with no prior app-building experience
📚 What I learned
- Prompting is powerful: With the right prompts, I could guide tools like ChatGPT and Bolt.new to generate structured, usable content
- APIs and deployment: I learned how to build with OpenAI, use Flask, and deploy projects with Render
- Adaptability: When one solution didn’t work (e.g., backend integration), I learned to pivot and find alternatives that still worked
🚀 What's next for HealthAidFinder: Your smart travel AI health assistant
- Add full integration between frontend and backend using a custom stack
- Offer multilingual support for global travelers
- Integrate insurance PDF parsing or API to verify real policies
- Add voice-based interaction for accessibility
- Expand location data to cover more healthcare options globally
HealthAidFinder is just the beginning — I hope to keep building on it to make healthcare navigation easier for everyone, everywhere.
Log in or sign up for Devpost to join the conversation.