Agentic systems with artificial intelligence (AI): potential uses in agriculture

Abstract

Talk covering the current context of artificial intelligence (2024 Nobel Prizes, the 2025 AI Index Report, AI maturity in Latin America) and its enabling factors: compute, data and algorithms. It reviews historical milestones in data analysis, the main neural network architectures (CNNs, RNN/LSTMs and Transformers) with agricultural applications, and the ChatGPT turning point. The core section introduces agentic AI systems (workflows vs. agents, architectures, ecosystems such as LangChain/LangGraph, Google ADK, Agno and n8n, the MCP, A2A and AG-UI protocols, and context engineering), shows in-house use cases (Cereza, Porkus, Nutribot, Feedibot) and closes with tools and learning resources to get started.