# E.ON AMS Agent

> Intelligent support in corporate logistics

- **Year:** 2025
- **URL:** https://fse-group.de/en/unsere-arbeit/e-on-ams-agent

For E.ON’s central logistics system we built an AI agent that detects anomalies, prioritises them and proposes concrete fixes – proactively, in context and in production. The outcome: 60% fewer tickets, 8 hours of relief per week – and more stability in day-to-day fulfilment.

When one system reflects every goods movement of a corporate group, stability is not a bonus but a prerequisite. E.ON AMS is such a system: it steers fulfilment, logistics and billing processes across the whole group – highly connected, business critical, complex.

But in distributed system landscapes anomalies are hard to spot. Error messages are not always real errors – and real problems disappear into the noise. For the team that meant manual research, hopping between systems, uncertainty.

Our answer: the E.ON AMS anomaly and support agent. It detects irregularities early, assesses them in context and proposes suitable fixes – including carrying them out directly in the system.

That changes support fundamentally: from reactive troubleshooting to proactive intervention. The team saves around 8 hours a week, support tickets dropped by 60% – with higher stability.

Delivery was demanding. Large volumes of data, complex dependencies, critical integrations – and at the same time the ambition not just to detect risks but to classify them meaningfully.

Interaction mattered to us in particular: the agent does not act abstractly but embedded in context – it offers actions, explains causes, but leaves the decision to a human.

### System landscape & integrations

  Security layer

  Oversight

  E.ON AMS
  Agent
  Anomaly & support

  Value
  Immediate detection
  instead of manual work
  Anomalies reported proactively

  LLMs
  Microsoft LLM Hub

  Logistics
  DHL · warehouse

  Monitoring
  Monitoring

  Human
  colleague
  Supervisor & approval

### E.ON AMS Agent – live demo

Simulation: a DHL shipment (ORD-2024-9912) has shown no scan update for 11 hours. The E.ON AMS Agent detects the anomaly automatically, retrieves shipment status, order data and the DHL damage report and establishes: package damaged. The agent proposes two actions — a replacement order via SAP (express) and a damage claim to DHL — which the staff member confirms with one click.

The result: relief in everyday work, greater safety – and an agent that takes responsibility without taking away control.

**Tags:** #AIAgents #AnomalyDetection #ERPIntegration #LogisticsSystems #SupportAutomation