Fundr

💭 The Vision Fundr transforms government funding from a bureaucratic puzzle into an AI-powered discovery experience. Our long-term goal is to help every individual, student, and nonprofit find the money that finds them.

💡 Inspiration Every year, billions of dollars in government grants and subsidies go unclaimed, not because people don’t need them, but because they don’t know they exist or can’t navigate the application maze. We wanted to change that. Our inspiration came from watching friends, students, and nonprofits struggle to find programs they actually qualified for. What if applying for help could feel as easy as swiping right?

✨ What It Does With Fundr, users sign up and fill in their profile details — like age, income, and residency — which can be updated anytime later. Once on the main page, it works just like Tinder for grants: our AI only shows funding opportunities that match your profile using our eligibility algorithm built with Gemini API. Swipe right to save a grant, which then appears in your Matches tab for review and application. You can also explore all available programs in the Search tab, and get instant help from our built-in AI chatbot that answers questions and explains eligibility. Fundr turns the complex world of funding into a simple, personalized swipe experience.

⚙️ How We Built It Fundr combines AI, data, and design empathy to match users with funding opportunities that truly fit their profiles. The FastAPI backend connects to Snowflake, which stores and cleans real subsidy datasets. Gemini AI reads subsidy descriptions, interprets eligibility rules, and compares them against each user’s profile using a semantic similarity approach. On the frontend, we built a Tinder-style swipe UI with React, Typescript, and CSS. The personalized ChatBot was built using Voiceflow.

🧩 Challenges We Faced Parsing messy government data, subsidy descriptions vary widely in structure and terminology. Managing API chaining between Gemini and Snowflake efficiently under time limits. Coordinating development between two people across frontend and backend branches in parallel.

🧠 What We Learned How to integrate large language models (Gemini) with structured data pipelines (Snowflake). The power of modular design: isolating services like gemini_service.py and pdf_service.py sped up collaboration.

🚀 What’s Next for Fundr Next for Fundr is fully automated applications. Once matched, users can submit directly with one tap. We’re also building smarter AI matching, a live tracking dashboard, and expanding to business and nonprofit funding across Canada. Our goal: make every eligible dollar apply itself, automatically.

Built With

• bash
• css
• expo.io
• fastapi
• gemini-api
• javascript
• json
• linux
• playwright
• pydantic
• python
• react
• react-native
• rest-api
• snowflake
• typescript
• uvicorn
• visual-studio-code
• voiceflow