Turning satellite datainto decisions.

TwinFarms is a precision agriculture platform that uses satellite imagery to show where fields need attention. The product combines maps, imagery, vegetation indices and operational data in a single interface used in production.

Geospatial workflows are heavy: imagery requests, index calculations and large geometries all have to be orchestrated without the interface feeling slow or confusing. Users need to draw and manage areas, compare dates and read results without understanding the pipeline behind them.

• Integrating Google Earth Engine workflows into product features.
• Building interactive mapping, layer and visualization features.
• Connecting the frontend, the Strapi backend and PostgreSQL data models.
• Deploying and iterating on Google Cloud Platform services.

• Keep expensive geospatial processing out of the request path where possible, so the interface stays responsive.
• Model areas, layers and results explicitly so the UI can render them without special cases.
• Treat map interaction as a first-class product surface, not a plugin bolted onto a dashboard.

• Layers and dates should be easy to compare without reading documentation.
• Failure states matter: imagery is not always available, and the interface has to say so clearly.
• Operational users work from the field; the product has to stay usable on smaller screens.

• Next.js
• React
• TypeScript
• Google Earth Engine
• Strapi
• PostgreSQL
• GCP
• Geospatial Data

Ongoing work on a live platform. The features I build ship to production users and are iterated on with the product team.