Our technology

We build tools that help you make more informed, efficient and scientific decisions.

Focusing on field perception and decision optimization, our technology collects, digitizes & analyzes vast amounts of data to help growers control & optimize their production and growing systems.

Convolutional (and transformer) neural networks for plant object detection

Get leaf-level perspective

We use machine learning to identify complex patterns in large, messy data sets. Using “deep learning” object detection, we can help you spot anomalies with leaf-level precision so you can treat crops for pests and disease sooner, with extremely high accuracy.

Multispectral Agro-signatures

Pinpoint problems early

You can judge the health of a plant from the electromagnetic radiation it reflects and emits. Using multispectral images captured by satellites, planes and drones, we develop multispectral agro-signatures that visualize crop vigor, water stress, nutrient deficiency, irrigation irregularities and more. This helps you identify issues early on, and to act before a problem causes crop damage.

State-action machines for growing policies

Optimize decision-making

Our technology can help you make optimal decisions to meet your objectives, all while saving your resources like water, fertilizer and chemicals. It dramatically reduces the amount of time and effort spent on daily decision making, and improves your efficiency across your operations.

For example, we all know irrigation is critical. Many growers overwater crops by default, because underwatering has such drastic repercussions. But because our irrigation recommendations are based on regularly monitored conditions, you can set the precise amount of irrigation you need, on a daily basis, to achieve optimal results.

Sensor-agnostic anomaly detection

Draw insights from multiple data sources

Our technology is hardware-agnostic, so it pairs seamlessly with a wide range of equipment such as Valley pivot systems. We process and analyze huge data sets from numerous sources to spot irregular patterns, regardless of the agro domain. This anomaly detection helps you get real insights and identify trends in areas such as pests, diseases, weeds and more. For example, by analyzing more comprehensive data you can easily tell if your production levels are meeting targets or not – and more importantly, what you can do about it.

Multispectral Agro-signatures

Pinpoint problems early

You can judge the health of a plant from the electromagnetic radiation it reflects and emits. Using multispectral images captured by satellites, planes and drones, we develop multispectral agro-signatures that visualize crop vigor, water stress, nutrient deficiency, irrigation irregularities and more. This helps you identify issues early on, and to act before a problem causes crop damage.

State-action machines for growing policies

Optimize decision-making

Our technology can help you make optimal decisions to meet your objectives, all while saving your resources like water, fertilizer and chemicals. It dramatically reduces the amount of time and effort spent on daily decision making, and improves your efficiency across your operations.

For example, we all know irrigation is critical. Many growers overwater crops by default, because underwatering has such drastic repercussions. But because our irrigation recommendations are based on regularly monitored conditions, you can set the precise amount of irrigation you need, on a daily basis, to achieve optimal results.

Sensor-agnostic anomaly detection

Draw insights from multiple data sources

Our technology is hardware-agnostic, so it pairs seamlessly with a wide range of equipment such as Valley pivot systems. We process and analyze huge data sets from numerous sources to spot irregular patterns, regardless of the agro domain. This anomaly detection helps you get real insights and identify trends in areas such as pests, diseases, weeds and more. For example, by analyzing more comprehensive data you can easily tell if your production levels are meeting targets or not – and more importantly, what you can do about it.