The professional conference “Use of Artificial Intelligence in Agriculture” showcased the latest trends and practical applications of AI in crop and livestock production. The event brought together leading Czech and international experts to discuss how modern technologies can enhance the efficiency, sustainability, and quality of agricultural production — as well as the broader agro-environmental and food security implications of AI.

Presentation by Karel Charvát on Behalf of PoliRuralPlus

Among the key speakers was Karel Charvát from Plan4all and the Czech Centre for Science and Society, who spoke on behalf of the PoliRuralPlus project.
His presentation, “Use of Language Models in Agricultural Management,” introduced how artificial intelligence and large language models (LLMs) can transform agricultural decision-making and knowledge systems.

From Utopia to Reality: The JackDaw System

Charvát presented the JackDaw (“Kavka”) system, a multi-agent platform that connects large language models with databases, spatial data, and analytical tools. The system automatically launches relevant analyses and provides users with clear, practical answers — turning what once seemed like technological utopia into a tangible solution for the agricultural sector.

Building the Foundation for AI in Agriculture

To enable the effective use of AI and language models in agriculture, several steps are necessary:

  • Adjust services and create unified catalogues of data and analytical tools.
  • Develop text libraries containing regulations, methodologies, and scientific papers.
  • Innovate map-based interaction, allowing users to query and visualize results directly in a spatial environment.
  • Run LLMs locally, ensuring privacy, independence, and compliance with EU data protection standards.

Rather than building one massive system, the aim is to develop customized, domain-specific AI solutions. Open-source language models make it possible to deploy local, privacy-preserving versions of tools like JackDaw, tailored to agricultural needs.

Looking Ahead: Federated Learning and Digital Sovereignty

A promising development path is Federated Learning, which allows AI models to train locally while sharing only improved parameters instead of raw data. This protects sensitive information and supports the EU’s ambitions for digital sovereignty, transparency, and sustainable transformation.

Shaping the Future Together

Speaking for PoliRuralPlus, Karel Charvát invited all interested stakeholders to join the effort to create next-generation tools for agricultural management. Together, we can shape how artificial intelligence supports everyday agricultural practice — building a smarter, safer, and more sustainable future for European agriculture.

📄 Download the full presentation (in Czech): Use of Language Models in Agricultural Management (PDF)

 👉 See the original event invitation (in Czech): https://www.ctpz.cz/udalost/konference-vyuziti-umele-inteligence-v-zemedelstvi-4-11-2025-praha-396

The conference was organized by the Czech Technology Platform for Agriculture in cooperation with the National Centre for Agricultural and Food Research, VÚMOP, Czech University of Life Sciences (CZU), Mendel University, University of South Bohemia, Leading Farmers CZ, World from Space, Farmtec a.s., and Plan4all z.s.

The event was held with the support of the Ministry of Agriculture of the Czech Republic.