SellinSeconds

The inspiration for SellinSeconds grew from the sheer chaos of current student commerce at UW-Madison. Right now, buying and selling happens in fragmented, unorganized WhatsApp groups and Discord servers where great deals are buried under hundreds of messages and "is this still available?" pings. As students ourselves, we noticed a heartbreaking pattern: incredible furniture and heavy essentials are often given away for free or sold for pennies, yet they still end up in a dumpster because the buyer has no way to move them. We realized that for most students, the barrier to a better living space isn't the price of a sofa, it’s the 10-block walk between Sellery Hall and an off-campus apartment while trying to carry it. We’ve always wished for a simple way to find a peer with a car or a dolly or willing to drive a U-Haul, and that longing for a built-in transport facility is exactly why we built this platform.

To turn this vision into reality, we engineered a robust backend that treats "helpers" as a primary pillar of the ecosystem rather than an afterthought. We built a specialized logistics layer using the Google Maps Distance Matrix and Geocoding APIs to resolve natural language searches like "near State Street" into precise coordinates. This allows the system to calculate proximity in real-time, ensuring that every match is geographically viable for a student on foot. On the seller’s side, we integrated an AI-first listing engine that uses LLMs to transform a quick photo and a messy chat prompt into a structured listing. This removes the friction of manual data entry, allowing sellers to post an item and get back to their studies in seconds.

One of our biggest technical hurdles was the "3-way chat" architecture. Traditional marketplaces are strictly 1:1, but we developed a custom orchestration layer that allows a standard buyer-seller negotiation to be seamlessly upgraded to include a verified student helper once their delivery quote is accepted. This preserves the full message history while pivoting the conversation toward logistics coordination.

Throughout this journey, we learned that the best technology is the kind that solves real-world physical problems. By utilizing LLMs to handle the cognitive load of categorization and Google’s geospatial tools to handle the physical complexity of campus navigation, we created a platform that feels intuitive rather than transactional. SellinSeconds isn't just a site for buying and selling, it is a logistics-integrated engine designed to make campus life more accessible and sustainable for everyone.

Built With

• claude
• fastapi
• gcp
• nextjs
• python
• supabase