How SOLaiR Works
Hybrid AI and geospatial assessment for solar installers.
SOLaiR combines Google Maps Platform, Google Solar API, computer vision models, climate datasets, and Hydro-Québec-specific financial modeling to move from a property address to a customer-ready solar proposal.
Book a Platform DemoHybrid Roof Assessment Engine
SOLaiR's roof assessment workflow combines four layers: Google Maps Platform and Google Solar API data for geospatial and solar context, SAM2-based roof segmentation with specialized prompting, YOLO-trained obstacle detection models, and geometric post-processing algorithms.
Google Maps Platform and Google Solar API provide satellite-derived solar and geospatial context.
SAM2-based prompting supports rooftop mask generation and boundary detection.
YOLO-trained models identify roof obstacles such as chimneys, vents, skylights, HVAC units, and other obstructions.
Geometric algorithms validate boundaries, estimate usable roof area, and prepare layouts for panel placement.
Solar Potential and Climate Modeling
SOLaiR uses Google Solar API data as a core input for solar potential analysis, then augments project assumptions with climate and production modeling, including PVGIS and NASA datasets where relevant.
Hydro-Québec Financial Modeling
The financial engine models electricity cost assumptions, purchase options, ROI, payback period, and 25-year savings projections for Quebec solar projects.
Google Cloud Roadmap
SOLaiR already uses Google Maps Platform and Google Solar API as core data infrastructure. As Claira scales, Google Cloud services such as Vertex AI, BigQuery, Cloud Run, and Cloud Storage are a natural fit for model inference, validation datasets, geospatial processing, and production deployment.
Google Cloud credits would help Claira scale AI inference, validation datasets, geospatial processing, and installer pilots without moving away from the Google solar data foundation already used by SOLaiR.