Unlisted

Inspiration

Our inspiration was one of our teammates personally getting "scammed" by Ion Austin #IHATEION. "I thought I was getting a really good deal for an apartment so close to campus, but I didn't consider all of the additional fees that quickly piled up."

Sherry

Many students don't have the time to properly investigate into every apartment and weigh every additional cost. Additionally, with the pressure of room's filling up, many students sign hastily. With Unlisted, we want to make this information transparent and easily accessible.

What it does

Unlisted helps students compare apartments beyond their advertised rent. Users can search for apartments manually or with their AI assistant to build a shortlist, explore floor plans, required fees, optional add-ons, and upfront charges. They can also explore more honest reviews from real tenants rather than inflated visitor reviews and view an interactive map of each apartment. Our AI assistant also helps students weigh the tradeoffs tradeoffs between cost, amenities, and resident experiences tailored to each individuals preferences.

How we built it

We built Unlisted with Next.js, React, TypeScript, and Tailwind CSS, using Google sign-in through Auth.js.

SerpAPI gathers public property information, Google reviews, and walking and biking distances.
Snowflake stores our structured apartment catalog and supports searches by floor-plan criteria, including bedrooms, bathrooms, and rent.
Meta Muse Spark interprets requests, extracts supported listing details, labels reviews, and explains apartment tradeoffs.
PostgreSQL and Prisma store saved apartments, conversations, and cached research.
MapLibre and OpenFreeMap power the interactive maps.

Challenges we ran into

Apartment websites present information differently, making it difficult to collect and compare consistently. Some list rent per person, others per apartment, and many leave utility costs unspecified. We had to preserve those differences and clearly flag missing information rather than assume a cost was zero. Researching multiple properties also made searches slow and used up API calls quickly. We introduced caching to reuse recent results and limited the number of properties researched per request to balance speed, coverage, and API usage.

Accomplishments that we're proud of

We’re proud of connecting multiple APIs and technologies to turn scattered apartment catalogs into a detailed, comprehensive dashboard. By bringing pricing, fees, floor plans, reviews, and commute information together, we made apartment comparisons easier to navigate. With this being our first hackathon, we are very proud of completing a working product within 24 hours.

What we learned

We learned how to take a product from ideation to a working prototype as a team, combining our ideas and coordinating frontend, backend, and design work. We also gained hands-on experience integrating APIs and technologies, turning inconsistent online information into structured data, and using AI to make that information easier to understand.

What's next for Unlisted

Our next priorities are broader property coverage apart from just Wampus apartments, fresher pricing, more complete fee information, and more consistent filtering across AI searches. We also want to improve review classification and make it easier for users to verify the sources behind each recommendation. Following this would be expanding to further campuses!