Project Jensen

Inspiration

Commercial building owners are currently squeezed by a "perfect storm": rising utility costs, stricter environmental regulations, and worsening water scarcity. Grundfos has the technology to mitigate these risks, but the sales process is often slowed by the difficulty of identifying high-potential sites. We were inspired to build a tool that automates this discovery, proving both environmental impact and business ROI at scale.ROI.

What it does

Project Jensen is an automated prospecting engine for Grundfos that uses computer vision on satellite imagery to detect commercial rooftops over 100,000 sq ft across the US. For every detected building, it calculates a Viability Score based on roof catchment area, local rainfall, water utility costs, and ESG commitments, giving Grundfos sales teams an instant, data-driven map of their highest-ROI targets for rainwater harvesting system deployment.

Data Ingestion: Aggregating diverse datasets, from municipal utility portals to corporate 10-K filings, required complex data normalization.

Resolution: Generating accurate Confidence Scores for specific rooftop equipment via satellite imagery requires high-precision visual analysis

Challenges we ran into

CV: As first-time hackathoners, deciding which CV approach to use was our biggest hurdle — we had never worked with computer vision before and had to quickly evaluate multiple technologies before finding the right solution.

Market Fragmentation: The rainwater management industry is constrained by a lack of standardization and fragmented local data, making a universal engine difficult to build.

Data Ingestion: Aggregating diverse datasets, from municipal utility portals to corporate 10-K filings, required complex data normalization.

Resolution: Generating accurate Confidence Scores for specific rooftop equipment via satellite imagery requires high-precision visual analysis

What we learned

We learned that water is the next major frontier for financial ROI in sustainability. For instance, a large commercial roof in a high-utility-cost state like California or a water-scarce region like Arizona can yield millions of gallons, significantly reducing operational expenses. Building this tool taught us that the "possibility in every drop" is best realized when physical satellite data is paired with real-world financial drivers like payback periods and tax rebates.

Built With

• fastapi
• python