CASE STUDY

AI support for science-based decisions

Ekan has been entrusted with developing AI-based working methods to make decisions easier with support from the current state of scientific knowledge. The customer uses the support as part of its climate protection work.

The client: The customer is a technical/infrastructure management company in one of Sweden's metropolitan municipalities.

The client's challenge: One of the most important issues for municipalities to work on in the medium term is technical and infrastructural climate protection. One challenge is that it is both a comprehensive and complex scientific field that is constantly developing. The customer needed support to be able to continuously compile the current state of research, find relevant data and based on this, develop rational proposals for decisions.

Our solution: Ekan helped the municipality by reformulating their needs into tasks that a language model is adapted to perform. The next step was to develop a simple so-called RAG model* based on available tools. This allowed us to test and develop both the requirements, the working methods and the solution together with the customer. We then connected specialist knowledge in AI and data science to jointly develop a business-adapted and information-secure support with high scalability. *Simply put, RAG is about the language model being able to refer to sources.

The result: With the help of AI support and new working methods, the operations can investigate, discuss and make decisions on issues on a scientific basis. The administration and the municipality have also received a concrete example of how and for what AI can be used in the various challenges of the operations. Human resources in the administration can now focus on knowledge-based situation analysis instead of continuously compiling, evaluating and cataloguing new knowledge. With better decision quality, the accuracy of the enormous investments being made to climate-proof society increases.