2nd research year for our FFG project "Virtual Advisor"

With the "Virtual Advisor", we give banks and insurance companies the opportunity to present suitable products to their customers digitally, specifically and automatically.

The "Virtual Advisor" makes use of a widespread phenomenon: data is collected almost everywhere and at all times. In most cases, however, that's as far as it goes. The data gathers dust unused on servers or is (automatically) deleted after a certain period of time.

Our research project aims to make data usable and extract usable information from it. Artificial intelligence uses this data as a basis for recommendations. "We are creating a service in this area that benefits both banks and insurance companies as well as their customers," says Wolfgang Stockner, CEO of bluesource.

Payment apps from banks are the ideal habitat for our "Virtual Advisor". It should be integrated there. Its task is to help banks and insurance companies to address customers in a more targeted and effective way. Be it for support, with interesting offers or personalized suggestions. Everything is clear and simple for users in the foreground. In the background, however, a range of state-of-the-art technologies are used.

The "Virtual Advisor" developed by us in collaboration with the University of Applied Sciences Upper Austria (Fachhochschule Oberösterreich - FH OÖ) is essentially a recommender system that is based on artificial intelligence (AI) and uses algorithms. "As with all data and AI projects, the biggest hurdle is making the various data usable and integrating it," explains Thomas Otzasek, Chief Data Officer (CDO) at bluesource. "We have built up an extensive data lake for this purpose. This is technically state-of-the-art." All available data has already been integrated. A gamification approach was tested to offer users even more incentive to participate. "We have thus created a framework for testing the algorithms under development," says Otzasek.

We are currently researching these algorithms together with the University of Applied Sciences Upper Austria - specifically with Prof. Ulrich Bodenhofer. He is a proven expert in this field, as evidenced by his numerous papers.

The "Virtual Advisor" also incorporated findings from two master's theses that were implemented at our company as part of this research project. A data-centric AI approach was pursued in the work in order to build up a good data source for the algorithms. Two further master's theses, which could also provide insights for the research project, have been announced.

As a federal funding agency, the Austrian Research Promotion Agency Österreichische Forschungsförderungsgesellschaft FFG provides reliable and targeted support for research, development and innovation in Austria. Our project is funded via the FFG's basic program. It is the first of its kind whose machine learning algorithm uses data from sources that were previously not available to banking and insurance service providers in this form.

You can find out more about the start of our research project and the first year of research here.