# Bayer's AI agents diagnose crop-research workflow failures

> Bayer reports agents investigating research failures; its support result lacks a published sample or audit.

By BIGCHANGE.AI Editorial

Published: 2026-10-25T04:36:28.730Z
Updated: 2026-10-25T04:36:28.730Z
Canonical: https://bigchange.ai/blog/bayer-ai-agents-crop-research-workflow-diagnosis

![Conceptual charcoal illustration of generic seedling trays on a greenhouse bench beside a wall-mounted terminal with an abstract orange interruption in unlabelled traces.](https://bigchange.ai/api/media/file/bayer-workflow-diagnosis-greenhouse-hero-v1.png)
AI-generated conceptual illustration by BIG CHANGE; no actual Bayer facility, interface, diagnostic result or crop outcome is depicted.

Bayer says its agents diagnose failed machine-learning workflows in the Connected Decision Engine (CDE). An [October 23 AWS account](https://aws.amazon.com/blogs/machine-learning/how-bayer-crop-science-built-an-agentic-decision-engine-on-amazon-bedrock-agentcore-with-aws-unified-operations/) reports gated use by over 100 staff.

## The big change

- **What changed:** Agents answer questions about CDE records; CDE runs the analytics.
- **Why it matters:** Bayer reports about 80% of first-phase questions resolved without engineer help; no sample or independent accuracy check is provided.
- **What to watch:** For research-operations teams, the practical decision is which records an agent may read and how to tell whether its answer resolved the user's question.

## The diagnostic route

Bayer describes a Microsoft Entra ID proxy, a LangGraph supervisor and 22 specialists. A troubleshooting specialist reads CloudWatch telemetry and CDE status through AgentCore Gateway with read-only tools.

[AWS's OpenAPI target documentation](https://docs.aws.amazon.com/bedrock-agentcore/latest/devguide/gateway-schema-openapi.html) explains how Gateway turns a described REST operation into an agent-callable tool. Its [authorization guide](https://docs.aws.amazon.com/bedrock-agentcore/latest/devguide/gateway-inbound-auth.html) separates access to the gateway from authorization at a downstream service. Bayer's permissions and diagnostic accuracy remain unverified.

## Assessing the reported result

Bayer reports diagnosis falling from over an hour to under 10 minutes, without a measurement method. No crop or model outcome is measured.

[AgentCore's observability documentation](https://docs.aws.amazon.com/bedrock-agentcore/latest/devguide/observability.html) distinguishes built-in operational metrics from additional, application-specific telemetry. A research team evaluating a similar interface needs a task outcome measure: a successful software call by itself cannot establish that the user received the right explanation.

## Sources

- [Bayer and AWS production account](https://aws.amazon.com/blogs/machine-learning/how-bayer-crop-science-built-an-agentic-decision-engine-on-amazon-bedrock-agentcore-with-aws-unified-operations/) — Coauthored case account; measures lack disclosed samples and audit.
- [AWS AgentCore Gateway: OpenAPI schema targets](https://docs.aws.amazon.com/bedrock-agentcore/latest/devguide/gateway-schema-openapi.html) — Explains translation of OpenAPI-described REST operations into MCP tools; general service documentation, not a test of Bayer's deployment.
- [AWS AgentCore Gateway: inbound authorization](https://docs.aws.amazon.com/bedrock-agentcore/latest/devguide/gateway-inbound-auth.html) — Distinguishes Gateway caller access and delegated downstream authorization. It does not disclose Bayer's policies or establish their effectiveness.
- [AWS AgentCore Observability overview](https://docs.aws.amazon.com/bedrock-agentcore/latest/devguide/observability.html) — Describes built-in and custom telemetry and CloudWatch traces; does not independently verify Bayer's task-success measures.