# Service requests solved before a human steps in

> Relief in customer care

- **Year:** 2026
- **URL:** https://fse-group.de/en/unsere-arbeit/customer-care-agent

For SENEC, FSE built a customer care agent that analyses service requests and pulls together information from CRM, ERP (including logistics), ticket system, knowledge base and the technical status system. After six weeks the first-time-fix rate was above 50 percent, and more than 45 percent of service cases now run autonomously.

SENEC develops and operates battery storage and energy solutions for private households. With every storage unit installed, the number of service requests grows. Many of them sound simple but require real work: before the customer care team can answer, it has to gather information from several systems.

For a single request, staff often opened the CRM, the ERP, the ticket system, the knowledge base and the system status view of the installation. The effort was not in accessing the data. It was in searching and assembling it. Request volume rose, handling times with it, and the backlog grew.

### What the agent does

Together with SENEC, FSE built a customer care agent. It reads incoming requests, recognises the matter at hand and fetches the right information across system boundaries: customer data from the CRM, order and delivery status from the ERP, open cases from the ticket system, instructions from the knowledge base and the live state of the installation from the status system.

Simple cases it answers itself by email. If a case is unclear or needs judgement, it prepares the answer and hands over. That creates a hybrid division of labour: the agent takes the research and the recurring matters, the team keeps the complex cases and the personal advice.

### System landscape & integrations

  Security layer

  Oversight

  Customer care
  Agent
  AI agent

  Value
  1st-level support
  handled autonomously
  Automatic email reply

  CRM
  Customer data

  Installation status
  Technical live status

  Ticket system
  Open tickets

  Knowledge base
  Docs & guides

  ERP
  Order & logistics

  Human
  colleague
  Escalation & advice

### Customer care agent – live demo

Animation: a SENEC end customer reports a red light ring on their storage unit (SENEC.Home) by email. The customer care agent reads the request, searches the knowledge base (match: communication fault after firmware rollout, fix: restart via the mein-senec app) and checks the installation status (online, firmware updated yesterday, warranty active). It then answers the request autonomously by email with the fix. The customer confirms: the restart helped.

### Results after a few weeks

The goal for the first year was a first-time-fix rate of 50 percent. It was reached after six weeks. The plan was to let around 30 percent of service cases run autonomously. After eight weeks it was more than 45 percent.

The backlog went down and the team gained time for demanding matters. The CRM team highlighted how accurately the agent classifies requests. In several cases it was closer to the right answer than the previous automation.

Connecting the systems you already have buys more service quality than any additional tool.

**Tags:** #CustomerCare #AIAgent #ServiceAutomation #Energy sector