Financial firms are at an inflection point, where incremental improvement is no longer enough. What’s needed now is a fundamental redesign of the operating model.
For years, the financial services industry has focused on efficiency, automation, and cost reduction. These priorities remain essential. But now, those capabilities are table stakes.
Today, leading firms need an operating model that can keep pace with accelerating market demands: higher volumes, longer trading hours, shorter settlement cycles, ongoing changes to Treasury clearing, and growing tokenization and asset digitization. The industry also faces rising expectations around transparency, control, and resilience.
Yet many operating models remain fragmented across platforms and workflows. In a more real-time environment, that fragmentation increases complexity, risk, and the burden of management oversight.
Granted, some firms have addressed these issues by automating individual processes using various technologies, most recently AI.
But firms need to move beyond isolated AI experiments toward a model built on a strong data foundation, API-first integration, a modular platform, and a common data model, all of which allow operations teams to shift from reactive, people-intensive processes to predictive, scalable operations.
In an always-on market, operating model design is no longer a back-office issue; it is a strategic capability.
What firms need now
At the SIFMA Ops conference in May, Quentin Limouzi, Global Head of Post Trade at Broadridge, hosted a panel titled “Next Generation Operations: Intelligent Automation for an Always-On Economy.” He was joined by RP Sandilya, Broadridge’s Chief Growth Officer of Wealth Management.
A clear theme emerged from the discussion: legacy models and piecemeal automation are no longer enough.
Markets are becoming more continuous, post-trade timelines are more compressed, and there is less tolerance for manual intervention. “There is no buffer overnight; there is little time to handle exceptions,” Limouzi said.
In the past, firms often solved operational problems by relying on the people performing the process or by making incremental changes to the legacy systems.
Now, firms need to look at how data flows through the organization to understand how the enterprise actually operates — how people work, where breakdowns occur, and where the root causes of problems truly lie.
“It’s not how do we automate more,” Limouzi said. “It’s really how do we build an operating model that can keep pace with where the markets are going.”
That is a challenge, Sandilya said. But it also presents a strategic opening.
“This is a pivotal moment for operations teams to think about how they approach the future,” he said. Firms have an opportunity to fundamentally redesign how data flows through the organization and how employees engage with each other.
Those moving forward need to ask five key questions, Sandilya said.
First, is your data reliable, accurate, and modernized? Second, is the automation that you’re putting in place dependable and repeatable? Third, can you explain the automation you put in place (can it pass an audit)? Fourth, is there a backup in place if there’s a failure? Fifth, if there is a failure, can you recover quickly and efficiently?
What a next-generation operating model looks like
Sandilya described next-generation operations as relying on a nimble, multifunctional workforce able to work across functions. In his view, those teams are also extraordinarily collaborative, sharing best practices in ways that raise performance and productivity. Just as important, staff must be well educated about the data that underpins the business and be able to ask the right questions of their data.
A next-generation operating model will be more unified and less dependent on fragmented manual work. It will not eliminate people, but it will change how they work and where they add value.
Limouzi framed the goal as a shift from reactive to predictive operations, stressing that “predictive” is the key word.
Operations teams also need to be able to direct and control their tools.
As he put it, the future operator works in an environment that allows them to control their “AI minions.”
This is not just theory. Sandilya noted that Broadridge has also been applying these ideas across its own “1,000-strong” operations organization, testing use cases that go beyond AI agents, email monitoring, and customer-inquiry automation to functions such as account onboarding and maintenance, fails management, break resolution, reconciliations, alternative investments, and unstructured data processing.
The warning
At the same time, the discussion made clear that applying AI to existing processes without fixing underlying problems can create new risks.
Sandilya offered perhaps the panel’s sharpest warning: “The single biggest problem you have is accidentally scaling inefficiency.”
That risk is especially relevant now because financial institutions have a long history of adopting new technology pragmatically, layering things like business rules and workflow tools on top of legacy environments to solve immediate issues.
“The best way to implement a use case and get something quickly to market was to layer on top of legacy technology,” he said.
That approach worked in the short term, but it also helped create the fragmented landscape firms are grappling with today.
Sandilya said firms now have a rare chance to fundamentally rethink the design.
But he stressed that intelligent automation brings a different level of responsibility than earlier generations of automation. Firms must account for privacy, governance, resilience, and explainability.
Those are not side issues. They are central to whether AI-enabled operations can be trusted in a highly regulated, exception-sensitive environment.
The payoff
If firms get the redesign right, the payoff is significant: better scale, earlier intervention, stronger controls, and a more resilient operating posture.
Limouzi repeatedly returned to the need to “scale without just adding people” and to “improve control while increasing speed.”
In an always-on market, firms cannot keep meeting new demands by expanding headcount in lockstep with complexity. They need operations that can absorb more volume and more volatility without becoming more fragile.
Sandilya pointed to where that value begins: with the ability to use data to understand where problems originate and to redesign work accordingly.
He cited early results from Broadridge’s own work, saying the company is seeing benefits on the order of 30% to 40% in productivity and error reduction across some use cases.
While outcomes will vary by function and firm, the broader point is clear: intelligent automation can create meaningful value when it is grounded in the right operating architecture.