AI is often mentioned in very generic terms. How is it actually integrated into SBP Regulatory Reporting?
AI isn’t a single, monolithic technology. We combine several approaches to address very specific customer needs into our SBP Regulatory Reporting solution.
- Symbolic AI: it’s the oldest form, but still absolutely essential. It relies on deterministic rules that come directly from regulations. These rules are traceable, auditable, and form the foundation of our regulatory decision-making.
- Statistical AI: this type of AI learns from data to support users, especially by detecting anomalies or breaks even before regulatory controls are applied.
- Generative and agent-based AI: this opens up new possibilities, such as easier access to documentation, orchestration of complex tasks (like creating reports), or assistance through copilots, always with human validation.
Can you give us some concrete examples of how this is used?
In practice, AI mainly helps us better support users in their day-to-day work. For example, we have an embedded chatbot that guides customers in using the product. If a client encounters an error, they can ask a question and immediately get an explanation of the error code, the associated regulatory rule, or a link to the relevant documentation. When you consider that the documentation can run into hundreds of pages or thousands of rules, the time saved and peace of mind is huge.
Another major use case is early anomaly detection. If a bank’s information system evolves (a new activity, a structural change), everything might still pass regulatory checks… but cause issues at the final reporting stage. With AI, we can flag unusual changes much earlier, based on the client’s historical data. The goal isn’t to replace rules, but to assist users, anticipate issues, and further secure the reporting process.
As R&D Manager for Risk & Reporting, you’re right at the heart of SBP RR’s evolution. What does your role involve?
Today, my role revolves around two main areas. On the AI side, I lead cross-domain initiatives. That means driving change, supporting development and business teams, and showing how these tools can be applied to real-life use cases to improve productivity. Innovation has to come from the field as well. I also oversee research into future AI features, including our first SBS CIFRE PhD on anomaly detection, and I carry out technology watch activities in close collaboration with our architects. Alongside this, I manage the team responsible for SBP Regulatory Reporting performance: defining the testing strategy, automating tests, analyzing results, and prioritizing optimization efforts.
What’s the biggest AI-related challenge you encounter as an R&D manager?
The real challenge with AI isn’t building a chatbot or a standalone tool, everyone can do that today. The real challenge is integrating AI into an existing product, across the entire lifecycle, in a way that’s industrialized, maintainable as models evolve, delivers relevant answers, and comes with a clear economic model. That’s the real issue.
After 10 years at SBS, how have your responsibilities evolved?
I started out as a tech lead on the launch of Anacrédit, the very first reporting solution at such a granular level: a response to the growing demand for increasingly detailed regulatory information. In 2020, I joined the SBP Regulatory Reporting journey, a product built on big data technologies and a cloud architecture. I led feasibility studies, coordinated up to six teams (in France and India), and supported the product all the way to its first production release in 2023. Since 2024, my role has become more transversal, with a strong focus on product performance, specific studies, and the gradual integration of AI into both our offering and our processes.
Which project are you most proud of?
Without hesitation, SBP Regulatory Reporting. It’s the project that fascinated me the most. We started out as a team of three: the technical architect, the functional architect, and myself as manager. We watched the product grow: more developers, more configuration experts, more teams. What’s most rewarding is seeing it now running successfully at customer sites and representing a real renewal for our product range. Being part of that journey from day one is what I’m most proud of.
What does it feel like to see the product you work on being used by customers?
One great thing about SBS is that once a product is developed, it reaches customers quite quickly. Since I started, I’ve already had the chance to contribute to the launch of two new products (Anacredit and SBP Regulatory Reporting). Before that, I worked more on long-term R&D, where you don’t always see the immediate impact. Here, you immediately see that your work matters. It’s rewarding, but also challenging, especially when it comes to performance, because customer satisfaction really has to be there.
A memorable anecdote?
A few years ago, before SBP Regulatory Reporting went live, I organized a boat trip on Lake Annecy with the team to celebrate our work and thank everyone for their efforts. The weather was perfect, everyone loved it, and it was a really nice moment to mark everything we’d achieved before the launch.
Would you like to join the SBS (ex-Sopra Banking Software) adventure? Discover all our job offers on our Careers page by clicking on this link.
You may also be interested in other talent portraits from our insights catalog.