Inspiration
The Restb.ai challenge inspired us to treat house hunting like a battle for the Iron Throne. We realized finding a home isn't just about budget, but identity. We wanted to solve the "chaos of data" by mapping specific Game of Thrones archetypes (like Cersei’s need for privacy or Jon Snow’s need for community) to real-world Los Angeles neighborhoods.
What it does
Joc de Barris is a strategic real estate advisor. It profiles users based on GoT characters and recommends their ideal "kingdom" in LA. Unlike simple filters, it features a Justification Engine that provides a data-driven explanation for why a specific neighborhood was chosen , visualizing everything on an interactive "Painted Table" map.
How we built it
We forged the frontend using React, Vite, and Leaflet for the interactive map. For data, we mined the Overpass API (OpenStreetMap) to extract specific nodes like schools, parks, and walls. The core logic is a custom "Reverse Tree" algorithm that normalizes these metrics and scores neighborhoods based on the specific weights of the user's archetype.
Challenges we ran into
"Data is dark and full of errors." Our biggest challenge was cleaning raw open data (handling nulls and outliers) to ensure accurate recommendations. We also had to design a flexible architecture capable of adapting instantly to the "Secret Client" requirement without hardcoding rules.
Accomplishments that we're proud of
We are proud of our Justification Engine. Instead of a "black box" AI, our system transparently explains the decision (e.g., "This area has 98% privacy for Cersei"). We also successfully successfully gamified complex data analysis into an immersive, cinematic UI.
What we learned
We learned the intricacies of geospatial engineering and how to query complex relations using Overpass. We also discovered that wrapping technical data in a strong narrative (GoT) makes data analysis much more intuitive and engaging for the user.
What's next for HackEPS-25 - Bytes 'N Roses
We plan to integrate a robust Python/Flask backend for heavy data processing and implement machine learning models to predict neighborhood trends—forecasting where "Winter" (gentrification or economic shifts) might hit next.
Log in or sign up for Devpost to join the conversation.