A climate-tech startup's flood-risk model took over two hours per property and needed manual tuning. Trexin built an automated, scalable MVP on AWS in under 60 days, cutting analysis from two-plus hours to under five seconds per parcel.
A startup had built a model to predict flood risk for a parcel of land over time and wanted to commercialize it to help people, businesses, and governments prepare for coastal flooding, storm surge, and sea-level rise. But the process needed heavy manual intervention and took over two hours to analyze a single property. The CEO asked Trexin to build a fully automated MVP, a working prototype in 60 days, to present to Miami-Dade city leaders at a conference.
The model drew on huge geospatial sources, LiDAR (nearly 2 billion data points for a single county), tidal gauges, and hurricane storm-surge (SLOSH) data, so we needed a scalable, cost-aware big-data solution. We built on AWS infrastructure with PostgreSQL extended by PostGIS, packaged in Docker containers for easy development and deployment. The hard part was emulating a data scientist’s judgment for cases outside the formula’s parameters; we solved it through weekly working sessions comparing automated vs. manual output and iterating the model until it ran fully automated.
In under 60 days, we delivered a scalable prototype that scores a parcel’s flood risk with no user intervention, and across a portfolio of addresses, cutting calculation time from over two hours to under five seconds per parcel. The MVP went on to become a publicly accessible commercial product.
We automated expert judgment and built the architecture to scale it, turning a slow model into a shippable product.
case study
case study
event
Tell us what you're trying to do. A senior practitioner will read it.
Talk to us