AutoMatch

LandingPage
CoreFunctionality

Inspiration

This project started from a real pain point we all faced while searching for second-hand cars. The process was exhausting endless scrolling across multiple marketplaces, poor filters, and overwhelming choices that made it hard to find the right match. We wanted to reimagine the experience: what if finding a car could be as fun and intuitive as swiping on Tinder? That question sparked the creation of AutoMatch.

What it does

AutoMatch lets users describe their dream ride in natural language, for example, “a sporty sedan under $25k” or “an electric SUV with 300+ miles range.” Our system scrapes data from major second-hand marketplaces, curates the most relevant cars, and presents them in a sleek swipeable interface. Swipe right to save, left to skip. It’s that simple.

How we built it

• Frontend: Next.js + React with Tailwind CSS and Framer Motion for animations and MapBox for Accurate Map Tracking including GeoLocator and GeoJSON. • 3D Experience: Three.js & React Three Fiber for the landing page with interactive car visuals. • Backend: Node.js with Express. • Database: PostgreSQL for user data and favorites. • Data: Web scraping scripts that pull from multiple used-car marketplaces. • AI Layer: NLP-based matching that interprets user prompts and returns relevant cars.

Challenges we ran into

• Parsing inconsistent data across different marketplaces. • Optimizing performance while rendering 3D models and animations. • Making the swipe deck responsive and smooth across devices. • Balancing AI prompt-matching accuracy with speed.

Accomplishments that we're proud of

• Built a fully working Tinder-for-cars prototype in limited time. • Designed an engaging landing page with interactive 3D animations. • Implemented smooth swiping with a polished UI/UX. • Successfully integrated prompt-based AI search with real scraped data.

What we learned

• How to integrate 3D visuals (Three.js) with a modern web app. • The importance of clean, normalized data for AI matching. • How to design user flows that make complex tasks feel fun and simple.

What's next for AutoMatch

• Expanding marketplace integrations for broader coverage. • Adding advanced filters (budget, mileage, body style). • Improving AI matching with embeddings and semantic search. • Launching a mobile app for swiping on the go.