AgriLLM is a collaborative AI initiative leveraging advanced language models to expand equitable access to agricultural knowledge worldwide. It is developing a domain-specific large language model (LLM) for agriculture, benchmarked against leading systems and to be released open-source as a public good - bridging research, policy, and field-level impact.Â
Established by the International Affairs Office at the Presidential Court of the UAE, with AI71 as the technology partner, AgriLLM is co-led by over 15 global partners including CGIAR, EMBRAPA, the Gates Foundation, FAO, IFAD, the World Bank, and ECHO, alongside universities and research centers.Â
Â
This COP30 session will offer an early look at the initiative’s journey so far - what AgriLLM is, how it was designed, and its grounding in scientific knowledge and local realities - as well as a preview of what’s next ahead of its global launch in December 2025.
Â
Event objectivesÂ
• Introduce AgriLLM’s vision, design principles, and collaborative architecture.Â
• Share insights from the development journey: data partnerships, user personas, and localization methods.Â
• Highlight how AI can enhance extension services and agricultural policy through inclusive design.Â
• Engage partners and stakeholders to shape the December 2025 launch and future roadmap.Â
Â
Â
Key outcomes
• Broader awareness of AgriLLM’s mission, technical foundation, and partner ecosystem.Â
• Early alignment on co-development priorities and governance principles.Â
• Identification of opportunities for pilot collaboration post-launch.Â
• Shared understanding of the initiative’s role in advancing responsible agricultural AI globally.Â
Presentation on AgriLLM’s development journey — how it began, what problem it addresses, how it’s being co-developed with over 15 partners, and how it leverages Falcon LLM to create an open, domain-specific model for agriculture. Includes a snapshot of user personas (farmers, extension agents, policymakers) and design principles emphasizing inclusivity and local context.Â
Â
Speaker:Â Mamoun Alaoui, Product Manager/ Technical Lead, A171
Reflection on how AI tools like AgriLLM can complement existing agricultural research, local extension networks, and field practices — ensuring solutions are regionally adapted and responsive to real-world agricultural challenges.Â
Â
Speaker: Silvia Massruhá, President / Head of Digital Agriculture,EMBRAPA
©2025 | Privacy Policy | Contact Us | COP27 | COP28 | COP29