There’s a moment in the 2016 Go match between AlphaGo and Lee Sedol that people who were watching still talk about. Move 37. A placement so unexpected, so seemingly wrong, that the commentators assumed it was a mistake. It wasn’t. It was the move that changed everything – not just that game, but how we understood what intelligence could do.

I’ve been thinking about that move a lot lately.

I was in San Francisco last month at the Databricks Data and AI Summit. Walking those floors, sitting with clients and partners, listening to where this industry is heading, I kept coming back to the same feeling: most of the people in that room have had their Move 37 moment. The point where agentic AI stopped being an experiment on a roadmap and became something more fundamental. A shift in how we think about what’s possible, and what’s expected.

What struck me most wasn’t the technology on show, though there was plenty of it. It was the conversations. Client after client, across geographies and strategies, describing the same inflection point in different words. The bar has moved. They’re not asking whether AI belongs in capital markets workflows anymore. They’re asking why their data isn’t ready for it.

 

The bar has moved

For years, firms have been investing in cloud migration, analytics, and more systematic ways of working. The arrival of agentic AI has raised the standard almost overnight.

Clients aren’t asking for data access anymore. They’re asking for AI-ready data: validated, governed, fit to use in workflows where the cost of being wrong is significant. They’re asking harder questions about lineage, auditability, sovereignty, and operational control.

When an AI-enabled workflow is routing an exception, supporting a portfolio decision, or generating client-facing output, the data beneath it can’t be fragile. Governance can’t be an afterthought. The infrastructure can’t be operationally brittle.

Capital markets has always lived at the intersection of complexity and consequence, making AI readiness more than a matter of simply adopting new tools. It requires a data foundation that can withstand the demands of regulated, enterprise-scale workflows where every output must be defensible.

 

Why an Intelligence Fabric matters

The challenge is that most firms aren’t operating from a single, clean, unified environment.

Data sits across vendors, asset classes, geographies, and internal systems. The same instrument, position, or exposure can appear in different places with different meanings. Workflows still depend on manual reconciliation and institutional knowledge – habits that human-led processes could tolerate but automated ones cannot.

That fragmentation limits how confidently firms can act, explain, and embed AI into real workflows.

It’s why we have been building an Intelligence Fabric for Capital Markets: a connected, governed architecture that links data, workflows, analytics, and AI across the investment lifecycle. It’s the connective tissue that turns fragmented data into decision-grade intelligence that firms can trust to use, explain, and defend when it matters.

The truth is AI doesn’t remove the need for governance and domain context. It makes them more important.

 

Building with Databricks

This is why our partnership with Databricks matters, and why we’re announcing it now.

This isn’t merely about partnering with a new technology vendor. We’re building with an engineering organization that lets us focus entirely on what we do best: business knowledge, data management expertise, and a deep understanding of what capital markets clients actually need. Databricks handles the platform engineering at a level of scale and sophistication we couldn’t replicate alone, nor should we try to. That’s the right division of labor. It’s how you move fast without cutting corners.

What Rimes brings to this partnership is harder to replicate than infrastructure. It’s the accumulated knowledge of complex investment management challenges, the relationships built on years of operating in regulated environments, and the data management skills to turn AI ambition into AI-ready reality. Together, we’re building something that can operate at the pace our clients demand and the resilience levels their most critical workflows require.

For our clients, that combination accelerates how trusted, compliant investment data gets into the analytics and AI workflows they’re building right now.

 

Open architecture, because that’s the reality

Here’s another thing the Summit reinforced for me: the future of capital markets data won’t be defined by a single cloud, model, or platform.

Firms will continue to operate across multiple ecosystems. Different providers, tools, and operating models, depending on their size, asset mix, and regulatory environment. The winning architecture won’t ask firms to abandon that reality. It will help them create a more connected, governed foundation across it.

Our clients need to integrate into the data ecosystems they’ve already built and evolve at their own pace. Some are well along on cloud and AI. Others are still working through foundational issues. Our job is to meet firms where they are while helping them move toward where the industry is going.

 

From AI ambition to AI-ready reality

I’ll be honest about one more thing: There’s a particular feeling when you’ve been heads-down on something for months – refining agentic flows, working through the edge cases, pressure-testing the architecture – and then you walk into a room full of the most innovative minds in data and AI, and what you’re building is not just recognized but validated. That happened for us at the Summit. It matters. Not for the ego of it, but because it confirms we’re solving the right problems in the right way at the right moment.

Our clients need us to do more. More intelligence in the data. More resilience in the platform. More speed without sacrificing sovereignty. More partnership, not just provision.

Move 37 looked wrong until it didn’t. The firms that will lead in this next era of capital markets are the ones making those moves now. Not waiting for consensus, not running another pilot. Building. That’s what this is.