Sibos 2026: Smartstream's Thomas Steinborn on Real-Time Settlements and AI

Quick Summary

As global markets consolidate and move toward T+1 and even T+0 settlement cycles, the pressure on financial technology to perform in real time has never been higher. Thomas Steinborn of Smartstream explains at Sibos 2026 how the company is investing in cloud-native foundations and Artificial Intelligence to manage rising transaction volumes and increasingly complex exception management.

How is the shift to T+1 and T+0 impacting technology requirements?

The market is moving toward 24/7 operations with settlements shifting from T+2 to T+1 and, in some regions, T+0. Thomas Steinborn notes that while Smartstream has been enabling real-time reconciliations for decades, the current environment is defined by a significant rise in transaction volumes. This increase naturally leads to a higher number of exceptions and investigations. To address this, the company is focusing on platform stability and scalability, ensuring that infrastructure can scale out with workloads and scale down during quiet periods to manage the total cost of ownership for customers.

What are the core pillars of Smartstream's cloud and performance strategy?

Investment is heavily directed toward the foundations of the platform to ensure it operates effectively within a customer's private or public cloud environment, such as Azure or GCP. Key technical focuses include:

• Leveraging cloud-native footprints to ensure maximum scalability.
• Separating compute from storage to improve efficiency.
• Enhancing matching performance, which is critical when dealing with upwards of 50 million transactions where every second compounds.

How is AI being used to re-engineer financial processes?

Thomas Steinborn distinguishes between classic Machine Learning and Generative AI. Smartstream continues to innovate with its Affinity machine learning base, which focuses on explainable AI. This allows regulators and operators to understand the reasoning behind match rates. Internal benchmarks for their next-generation matching have shown performance improvements that can be significantly faster than traditional methods.

Regarding Generative AI, Thomas Steinborn suggests that the goal should not just be faster processes, which he compares to building a faster horse. Instead, the focus is on process re-engineering to create the equivalent of a rocket ship. This involves simplifying processes, institutionalising smart knowledge, and demonstrating a clear return on investment by reducing the time taken for investigations and data lookups.

What role do security and compliance play in modern release cycles?

Beyond AI, a major area of focus is maintaining security and compliance in an era of rapid change. Thomas Steinborn highlights the impact of DORA and the need for faster release cycles. While annual or six-monthly updates were once the norm, the industry is moving toward monthly or even bi-weekly shipping. This creates a challenge for maintaining SOCKS compliance and requires close partnership with customers to ensure they can verify and certify updates within these compressed timeframes without disrupting production.

To learn more about how these technologies are being implemented, visit: https://smart.stream/solutions/smart-agents/

FF NEWS TAKE:

The transition to T+1 is often discussed as a regulatory hurdle, but Thomas Steinborn highlights it as a fundamental catalyst for technological evolution. By focusing on the separation of compute and storage alongside explainable AI, Smartstream is addressing the dual needs of high-speed performance and regulatory transparency. The shift toward bi-weekly release cycles is particularly telling of the broader industry trend: fintech providers are no longer just software vendors but are becoming continuous integration partners who must balance rapid innovation with the rigid requirements of financial compliance.