Le Gouvernement du Grand-Duché du Luxembourg
Anne-Goujon-©mBrumat

Artificial intelligence: Public-private synergy

Early this year, BGL BNP Paribas launched an R&D project with the University of Luxembourg’s research centre SnT to develop an AI "metamodel" tool, with support from Luxinnovation.
BGL BNP Paribas was acknowledged for this initiative in a press statement as the first bank to propose an artificial intelligence project eligible for public co-financing under the R&D and innovation aid scheme. The three-year project of about €7.4 million over 3 years can be co-financed up to €2.95 million by the Ministry of the Economy.

“We’ve been working with AI tools for about seven years and with the help of the university, we’ve been able to put several AI solutions in production. Specifically, they helped us to address the challenge of integrating new technologies in legacy infrastructure, and I think it's one of the reasons why we’ve been ahead,” explains Anne Goujon, Director of Data Science Lab, BGL BNP Paribas. “Compared to what we can see in the market, we have been able to put 100% of our solution from the laboratory into production,” she asserts.

We do not plan to replicate what exists but to think in advance with our experts.
Anne Goujon, Director of Data Science Lab, BGL BNP Paribas

The present partnership with SnT is focused on creating an AI metamodel tool in order to effectively monitor these AI solutions in a fast changing financial, regulatory and economic setting. “We put in place a real-time monitoring, but this takes about 80% of the lab’s capacity which is highly impacting for us,” adds Ms Goujon. “We discussed with the university and agreed that we can't solve the challenge by adding more people but that we needed to work on monitoring in a different way to free up the time of our experts,” she continues.

The project further exemplifies how public-private partnerships can be leveraged to combine resources, expertise and perspectives to minimise risks linked to new innovations and to achieve a shared goal. “This partnership is really interesting because research and industry are two different worlds. The SnT researchers explained the risks of such a project because the topic was partly investigated in the academic environment, and not yet in industry. And they were the first ones to say we needed to talk to Luxinnovation first to see if such a project could be brought to the Ministry of the Economy,” she underlines.


Added value of Luxinnovation support


“Luxinnovation helped to analyse our project, they asked plenty of questions and I was really impressed by their professionalism, and they went really deep into the project. They helped us to structure it so that the roadmap is very clear and also clarified everything in terms of risk. They equally supported us to prepare the files needed to be able to discuss with the ministry. We had regular meetings with them on the application and the project itself,” she states.

We have been able to put 100% of our solution from the laboratory into production.
Anne Goujon, Director of Data Science Lab, BGL BNP Paribas

The metamodel development project is poised to adopt diverse AI technologies, initially focusing on machine learning, and later expanding to natural language processing and generative AI. The team is taking the first steps in its R&D endeavour and recently had a first meeting to discuss initial findings and insights. “We do not plan to replicate what exists but to think in advance with our experts,” asserts Ms Goujon, who sees the potential benefits of the project extending beyond the bank to the general financial sector and industry at large.


 

Need more information?

Contact Luxinnovation's Corporate R&D and Innovation Support department.
Contact us

Newsletter sign up

Read our privacy policy

What's new?

View all news View all news

Stay informed on our latest innovation activities.

Get regular updates via our monthly newsletters and periodic emails tailored to your preferences.

You may unsubscribe from your newsletter and email subscriptions at any time. For information on how to unsubscribe, as well as our privacy practices and commitment to protecting your privacy, please review our Privacy Notice.