City of Tampere

Solution

Capgemini and Sogeti helped the City of Tampere turn early generative-AI interest into a governed internal service architecture rather than isolated experiments. The work combined AI governance design, learning circles across city functions, and two hands-on pilots built on the city's own information sources: Ask HR and Ask About Procurement Contracts. Instead of opening models against all internal content at once, Tampere deliberately scoped the first deployment to HR information on the city intranet, using GPT-based capabilities to support dialogue and retrieval while reducing risk. Capgemini also designed a reusable architecture so additional use cases could be added without rebuilding the foundation each time. That made the project both a live knowledge-access implementation and a framework for scaling future municipal AI services under shared governance, training, and change-management practices.

Data flow

Targeted internal information such as HR content and procurement-contract material is scoped into the pilot architecture, where GPT-based assistants interpret employee questions and return answers through Ask HR or Ask About Procurement Contracts. Governance and learning practices wrap the flow so new use cases can be added without reopening all city data at once.