Michel helps you instantly

How do I get to the city lost and found office? When is the vehicle registration office open? How do I register a business? Employees of Hamburg's public authorities answer questions like these every day. For some time now, they have been supported by Michel. With the help of Artificial Intelligence, the chatbot has the right answer ready for many recurring inquiries, 24 hours a day, seven days a week.

Michel: "Hello! How can I help you?"
Caller: "Where can I apply for a new driver's license, please?"
Michel: "There are several locations in Hamburg for this service. Here are the addresses."
Caller: "Great, thank you very much."
Michel: "You're welcome, see you next time!"

Der Chatverlauf einer App ist auf dem Foto zu sehen.

At first glance, this is an information situation that is commonplace in many public authorities in Germany. However, in this case Michel is not an employee of the city administration, but a chatbot. Since April 2019, he has been taking care of the administrative concerns of Hamburg's citizens around the clock. The digital assistance system "Ask-the-Michel" not only provides reliable answers in nine different languages, but has also mastered manners such as the typical greeting and small talk. Michel currently conducts around 600 virtual dialogs a month, with a quarter of the inquiries coming in outside the telephone availability of the regular HamburgServices offering.

Intuitive conversation in context

To provide suitable answers to citizens' questions, Michel uses artificial intelligence. The system analyzes a question posed and attempts to derive the underlying intention based on a context model. To do this, it checks the input with regard to the (foreign) language used, corrects any spelling errors and calculates the probability of potential concerns. Michel reproduces his information in a natural dialog, answering follow-up questions in the relevant context. Unlike widespread information systems, the chatbot does not play out long, complex documents, but rather very specific hints depending on the concern. The context model ultimately enables an intuitive conversation or natural human-machine interaction.

As a knowledge base, the chatbot uses, among other things, the contents of the Hamburg Service Information System (HaSI). The usage data and user feedback help to continuously develop the system. However, Michel does not operate completely autonomously, but is actively trained - for example, in the case of questions to which it has not yet been able to provide satisfactory information. In such cases, an employee of the Hamburg city administration writes a suitable answer and decides whether the question belongs to an existing dialog category or whether a new category must be stored in the system. With these clearly defined question-answer pairs, the chatbot's internal models are continuously updated. Automated tests are also used to regularly check the correctness of the question-answer assignment.

More networked in the future

In the future, Michel, who is currently running a beta version in test mode, will expand his range of services. To this end, the city of Hamburg is determining the expectations and wishes of users from the live operation of the chatbot and deriving functionalities that have been missing up to now. In the future, for example, the chat will be made even more intelligent so that the user can move more quickly from the question to the optimal answer. In addition, forms and other online processes can be integrated directly into the dialog, so that information that has already been entered can be transferred without media discontinuity or multiple entries. Linked to the appointment system of the customer centers, available times can then also be booked directly and easily via Michel.

Application facts

Technology field
Natural Language Processing
Application industry
Other Services
Field of application
  • Miscellaneous
Value-added activity
Service/Customer Service
AI Developers

Freie und Hansestadt Hamburg


AI Development partner

Technische Universität Berlin