Could you explain the company’s activities?
Our ultimate goal is very simple: to help growers and agri-food companies make better decisions. To achieve this, we work on three pillars: artificial vision, both in the field and on processing lines; data consulting and engineering focused on structuring company information; and a private artificial intelligence agent, developed by us, which analyses those data and generates immediate value.
Let’s start with artificial vision. How does it work?
We use high-precision cameras installed on tractors that travel through the plots. They detect and count fruit in real time, enabling harvest forecasts with accuracy levels of up to 90%. This transforms farm management: logistics, sales, labour and harvest planning all start from a reliable and objective forecast.
We do the same on production and processing lines. We have clients in Spain that process cobs, for example: our cameras count, classify and identify product condition throughout the production flow. Each fruit or unit becomes a data point, and that data can be exploited. And what happens to the data? The agri-food sector is not usually well structured in this area.
Correct. In agri-food, data are often fragmented, duplicated or “dirty”. Our consulting work focuses first on homogenising, cleaning and organising that information.
Once this is done, we connect our artificial intelligence agent within the company’s private environment. The data never leave that environment; it is not a public chatbot. It works exclusively with the company’s own data, securely.
Do you have an AI agent developed by Agerpix?
Yes, we developed it ourselves. When we connect it to the system —ERP, CRM, farm logbook, Excel or any data source— it is able to understand queries in natural language: “How are we doing in production?”, “How many kilos went through the line last month?”, “Create a chart with sales from the last five years”. It responds in seconds, generates charts, tables and cross-analyses between departments.
For example, it can explain why sales of a product have fallen by combining production and warehouse data.
And, very importantly, access is restricted by role. An operator can consult operational data; an executive, strategic information. No “how much does my boss earn”.
Is this system accessible to employees from mobile devices?
Yes. It works from any device. Today, companies usually ask someone to prepare a dashboard or a weekly report. With our system, that time disappears: you request it via chat and receive it instantly. It is a major leap forward for management teams, because they no longer depend on an analyst or a department to understand operational performance. Intelligence flows across the chain, from the field to the office.
Within the fruit and vegetable sector, who are your clients?
We work with companies in Spain, France, Israel, Argentina, Peru and Chile, especially Chile. In artificial vision, we mainly collaborate in citrus and stone fruit: peaches, nectarines, apricots… We need large, visible fruit. If leaves cover the fruit, the camera cannot count it.
Berries and leafy crops (lettuce, baby leaf) do not make much sense for this system.
However, in data and AI we work with any agri-food company, regardless of crop. In Murcia, for example, we supply major table grape companies, where value lies both in data analytics and in bunch counting and plot mapping with estimated production.
Could you explain what a dendrometer is?
It is a device that measures the actual growth of fruit on the tree. We install it when the fruit is around 30 mm in diameter. From that point on, the dendrometer —which incorporates a chip— detects every millimetre of expansion and sends the information to a GPS and a mobile application.
The application draws the growth curve, projects sizes and trends, and alerts if something is wrong: lack of irrigation, water stress, nutritional problems…
For crops such as citrus or stone fruit, it is a decisive tool. These are our own patents. It is not an off-the-shelf accessory; it is technology developed by Agerpix.
Can the data generated by these devices create value across the entire chain?
Absolutely. With artificial vision we obtain counts and ripening patterns; with the dendrometer, physiological fruit growth curves; with the AI agent, actionable decisions.
This combination —field + processing + internal data— makes it possible to plan sales, adjust human resources, schedule harvests, reduce waste and negotiate better with customers and distributors.
