Agrinsight: an AgriTech platform for managing orchards remotely

The problem

Advice & Consulting designs intensive orchards and Italian-made storage systems for Russian-speaking markets. The real work is not drawing the installation: it is following it afterwards, when the orchard is thousands of kilometres away and the decisions that matter are taken by looking at last night's temperature and this morning's leaf wetness. Manual monitoring holds while there are few clients. Then it stops holding.

How we solved it

Agrinsight is the platform that holds the two halves of the problem together: the sensors installed in the orchard and the people who have to decide what to do. Data arrives from the field and is read from a phone or a computer, live or as history, with every sensor under the same gaze instead of one at a time.

A number, though, is not an instruction. So the platform does not stop at showing readings: it compares them with the optimal values for the crop and raises an alert when something moves outside the limits, saying when and how to act. That is the difference between knowing that humidity rose last night and knowing that this humidity, at this temperature, opens the window in which a disease takes hold.

Data that arrives once the disease has taken hold is not data: it is a news report.

Above the data sits the layer that makes all of it usable, and it is not technological: the agronomists. From the platform the grower talks directly to Advice & Consulting's technical team through an integrated chat, and every week receives the general orchard report, with the sensor data and the guidance of the people who read it. The scope covers the whole cycle: soil and water analysis, crop protection, fertilisation, irrigation, pre-harvest, harvest and cold storage.

The piece we are proudest of is the one you only notice when it is missing. The agronomists' app is in Flutter with an offline-first approach, because in the middle of an orchard there is no signal: you record what you see, and synchronisation happens when the network comes back. An app that needs a connection to save an observation is an app the agronomist stops opening.

Underneath, the architecture is cloud-native and serverless on AWS, with a Node.js, TypeScript and Fastify backend and the API layer separated from the workers, so the two scale independently. Sensor data ingestion is orchestrated at hourly intervals, the web dashboard is in Angular, and synchronisation goes through AppSync and GraphQL. All of it delivered: front end, backend, DevOps and mobile.

The results

The result is measured two ways. On one side Advice & Consulting follows more clients at once at the same quality as before and with shorter response times, because tasks are assigned from the platform and the agronomic documentation is generated rather than written. On the other, out in the field, action becomes targeted: less water, fewer crop protection products, less fuel, and labour organised around what is actually needed rather than around the calendar.