AI-driven structuring of insurance data for Open Finance at F&P

Challenge

Insurance policies and terms are hard to extract and standardise data from – especially across companies. The traditional methods are manual, time-consuming and expensive, and they cannot scale to the EU's coming FiDA requirements. At the same time, F&P lacked a shared data model and the governance to support a sector-wide approach.

Solution

Capacit built an AI solution that reads insurance documents and turns them into structured data on a common data model designed to meet the FiDA requirements.

That is what the coming Open Finance rules ask for: Financial companies must be able to share customer data securely and in the same format, across company lines. With the model in place, data can move and that is the precondition for shared analysis, faster processes and new, more personalised products.

How AI adds value

The pipeline extracts the central insurance data fields with ~95 % precision, validated and the processing time per document has dropped sharply. All of it work that used to be done by hand, slowly and at a cost.

At the same time, the standardised data model provides a foundation for scaling across companies in the sector with significant efficiency gains and savings at sector level.