Bayer builds agentic decision engine on Amazon Bedrock AgentCore
In brief
- Bayer deployed ADE, a multi-agent platform on Amazon Bedrock AgentCore to query agricultural data in natural language.
- ADE resolves 80% of questions without technical intervention through specialized agents and a supervisor.
- The architecture uses LangGraph for orchestration and AWS Unified Operations for production readiness.
Bayer Crop Science developed the Agentic Decision Engine (ADE), a multi-agent platform on Amazon Bedrock AgentCore that lets breeders, analysts, and researchers query their agricultural data in natural language. According to the announcement, ADE builds on the existing Connected Decision Engine (CDE), which manages hundreds of ML model configurations for crop improvement.
ADE uses a supervisor agent to route questions to specialized agents, each querying relevant sources (APIs, metadata, dashboards) without users navigating the technical stack. In its initial phase, ADE resolves about 80% of questions without engineer intervention. The architecture consists of four layers: user interface, agent orchestration, specialist agents, and tools/data sources. Agents run in an isolated runtime on Amazon Bedrock AgentCore, with LangGraph orchestration for multi-step workflows.
Integration with AWS Unified Operations prepared ADE for production, with dedicated technical support. The article outlines five key design decisions, including identity management across technical boundaries and model compatibility with agent tools. The goal is to reduce workflow failure diagnosis time from over an hour to under 10 minutes.
What it changes for you
- AgricultureAgronomists and researchers gain direct access to real-time workflow diagnostics, cutting failure analysis time from over an hour to under 10 minutes.
This summary gives you the essentials. The full article, its figures and quotes are on Artificial Intelligence.
Read the original articleAI-generated summary of the source, checked by rules. How we write · Report an error