Remi 2.0: AI-Driven Roofing Solutions
Winning 1st Place at the Utah Developer Hackathon with an end-to-end roofing analysis and target marketing platform.
Remi 2.0: AI-Driven Roofing Solutions
Victory at the Utah Developer Hackathon was about more than just code; it was about solving a real-world friction point in the roofing industry using the latest in AI and Geospatial technology.
The Problem
Roofing quotes are traditionally manual, time-consuming, and prone to error. Marketing to potential clients with damaged roofs is equally inefficient.
Our Solution: Remi 2.0
We built two core pillars:
- AI Roof Analysis: Users input an address, and our system uses the Google Maps API to fetch satellite imagery. ML Vision models then analyze roof dimensions and solar energy potential to generate a detailed PDF quote.
- AI Target Marketing: A strategic tool that analyzes whole neighborhoods. Using Transformers and Vision models, we detect roof locations and evaluate their condition (e.g., potential damage) to generate personalized marketing materials automatically.
Technical Architecture
Impact & Recognition
Competing against some of Utah's top developers, our team (Jonghyuk Lee, Taehoon Kim, Yirang Lim, and myself) secured 1st Place. The judges included leaders from OpenAI, Google, and Y Combinator, who praised the immediate viability of the product.
I'm grateful to my mentors at Pattern® for their ongoing support and for encouraging this kind of bold innovation.
