AI chatbot deployed within 14 days (German only)

From emergency aid to AI agent

Meter readings, installment payments, billing: These are topics that sooner or later affect all TEAG customers, resulting in a high, constant volume of inquiries for customer service.

TEAG Thüringer Energie AG is an energy supplier based in Erfurt and has been a moinAI client since 2020. The company relies on an AI assistant as an immediate solution to relieve the burden on its customer service team.

It all started during the COVID-19 pandemic. With fewer staff available on the phones, there was a simultaneous surge in questions regarding installment payments and contracts. Back then, the chatbot went live in just 14 days, answering the most common questions around the clock.

Expectations have changed over time, and TEAG wanted issues to be resolved directly within the chat. Simply answering questions is no longer enough. This led to three goals:

1. Secure connection to core systems instead of general information
2. End-to-end conversations resolved directly in the chat with customers
3. A strong user experience without media discontinuity

One solution, five service processes

The new TEAG customer portal was launched in mid-2026, enabling the use of interfaces via an agentic AI platform. Anyone logged into the portal can get started immediately. All others authenticate themselves directly in the chat using the same logic they are already familiar with from the portal. Once logged in, dedicated AI agents take over. They read data from the system, verify it, and write changes back directly.

Today, five processes are available for tasks that users previously had to access via the customer portal or by contacting the service team:

• Adjust your advance payment
• Report your meter reading
• Check prices by postcode
• View your tariffs
• Set up a SEPA direct debit mandate

The process in detail: Adjusting installment payments

The AI agent first displays the current installment amount and the permitted range from the system. Then, one of three scenarios applies.

Installment payments were already one of the most frequent topics in the TEAG chat back in 2020. At that time, the chat could explain where customers could change their installment amount. Today, it changes it itself. Here is how it works:

1. Customer logs in and selects "Adjust installment payment," for example, via the quick-select menu in the chat.
2. The AI agent displays the status: current installment amount plus the minimum and maximum amounts from the system.
3. Customer provides the desired amount. Now the system decides:
   
   • Amount within range: The AI agent records it directly and confirms it in the chat.
   • Amount outside of range: The AI agent requests a correction.
   • Adjustment blocked, e.g., due to outstanding items: The system detects this immediately. Customers are informed of the reason.

Want to learn more?

In this case study, you will learn in detail how TEAG made the transition from a Q&A bot to AI agents and how the five processes are structured. Including all edge cases and the figures for the SAP integration. The architecture is modular, allowing for additional processes to be added at any time.

Learn, among other things:‍

• All five processes with conversation paths and rules for edge cases
• The security check for SEPA mandates step by step

‍

• A request as a diagram, from login to booking in the system
• Usage figures and scope of integration with webhooks and the TEAG API