SBS AI Foundation: Bringing banks from pilot to production

Banks have spent the last few years experimenting with artificial intelligence, but experimentation and real-world deployment are two very different things. Today, only around a third of banks have managed to scale AI into a core process. This gap between promising pilots and production-ready systems is exactly what SBS AI Foundation, the first product in the SBS AI portfolio, was built to close.

The Problem SBS AI Solves: Fragmented, Unready Data

According to William Remacle, head of SBS Data Platform, most banks run into the same wall. A pilot works beautifully in a demo: an AI assistant summarizes a corporate customer’s account activity, flags risk, and even suggests a relevant product. But when the bank tries to move that pilot into production, the cracks appear. Data is scattered across dozens of applications. Years of system migrations and bank mergers have left duplicate or incomplete customer records. In some cases, the same client exists twice in the system, and an AI tool has no way of knowing that.

“It’s not just about taking data and expose it to AI”

This is why SBS AI Foundation was built to help banks move from AI experiments to real-world deployment by making their data AI-ready and embedding secure, governed AI directly into their existing banking platforms. As Remacle puts it, “it’s not just about taking data and expose it to AI” ; the data has to be extracted, cleaned, and freed of redundancy before it’s useful. With 50 years of banking domain expertise, SBS has built a banking semantic layer and enterprise data model that generic AI vendors simply cannot replicate. As Remacle explains, outside vendors “are able to provide really good AI and models, but” they don’t know the bank’s products or “the banking domain either, as we do.”

Two Use Cases, Built for Immediate Impact

SBS AI Foundation launched with two use cases chosen for their immediate value: helping relationship managers better understand customer needs before every interaction and providing instant answers from the bank’s own governed data to support service, compliance, and operational teams.

Security and Governance Built In

“An AI answer alone is not enough.”

Because banking is one of the most regulated industries, SBS AI runs within the bank’s own environment, without external internet access, and respects existing user permissions. As Remacle notes, “An AI answer alone is not enough” ; every response must be traceable back to its source data, transformation history, and ownership, a governance requirement banks can’t compromise on.

A Progressive Rollout

“If a bank does not move, competitors will do.”

SBS AI Foundation is now available to select clients, with broader rollout planned for early 2027. Rather than a big-bang launch, SBS expects adoption to be gradual, expanding from these two use cases into more complex, agentic AI applications over time. For banks still on the fence, Remacle’s message is direct: “if a bank does not move, competitors will do,” and institutions that delay risk being “left behind.” SBS AI Foundation, he adds, is meant to be “a journey” (start small, prove value, then expand) rather than an all-or-nothing leap.