Avocado producer - dry matter quality prediction

The challenge

Harvest scheduling inefficiencies due to dry matter uncertainty led to quality issues and missed optimal harvest windows. Manual processes couldn't handle the complexity of plot-level heterogeneity, weather patterns, and terrain effects. Limited week-by-week supply visibility hampered labor and logistics planning across the operation.

What we built

We built an automated system that consolidates weather and field data, enriches it with agronomy-aware features like terrain and seasonality, and delivers weekly dry matter predictions. The solution features versioned reports, interactive dashboards, and a natural language Q&A interface for self-serve analytics.

What changed