Rimes Configures Capabilities into New Fabric to Scale AI Data Workflows

Rimes has built out its data management and investment intelligence offering to give clients a new unified suite of capabilities that can accelerate and scale their artificial intelligence-powered data and workflows.

“Frontier model capabilities are evolving rapidly. Our clients are focused on ensuring their data foundations and workflows can realize the benefits these models enable.” Mayadas told Data Management Insight. “All of those AI conversations inevitably pivot to a conversation about data and data foundations.”

The key aim of Rimes’ new integrated package is to serve clients with trusted data that they can deploy with confidence to their AI and agents. A recent A-Team Group webinar discussion among a panel of data leaders highlighted that highly accurate and complete datasets are of primary importance to modern data infrastructures.

Trust in Data

Rimes joins other vendors, including Lexis Nexis and GoldenSource, in responding to industry demand for trusted data as organisations transition towards more autonomous AI agents and place more analytical emphasis on AI models. They are seeking to fill a gap between clients’ AI ambitions and their capabilities, and to head off what Informatica from Salesforce has characterised as “trust paradox” that could pose operational and regulatory dangers if not addressed.

“We’ve seen a big shift in demand from the industry around trust, lineage, explainability and auditability in the underlying data infrastructure and data foundations,” said Mayadas. “What’s happened now is that the need for AI-driven workflows across the investment lifecycle is making what we’ve been doing table stakes. And it’s also asking vendors like us to lean into this AI opportunity and create scalability and resilience in our infrastructure so that agents can act and make decisions quickly on infrastructure that is trusted.”

Intelligence Fabric for Capital Markets is designed to empower clients’ decision-making capabilities and better position them to respond to shifts in the industry, said Mayadas.

Four Pillars

Although not a fabric in the specific data management sense of the term, the offering does deeply connect datasets on an architecture that’s built around four pillars, or layers.

• The first is the foundational layer of data management that Rimes has been providing to clients for three decades. The “trusted data network” integrates more than 1,000 data and technology providers, connects billions of data points and performs millions of validations per day.
• The next layer is what Rimes calls Connected Intelligence, a mastering strata that resolves entity disparities and brings datasets in line with each other. It also features a semantic overlay that enables better translation of intent to underlying data.
• The third layer provides the tooling that harnesses data within decision-grade workflows.
• And, the final element is what Rimes calls its “always-on operational control”, an observation tool that seeks and repairs irregularities within the data pipeline.

The fabric also benefits from functionality offered by Databricks’ Data Intelligence Platform, on which it is hosted. Clients can access the integrated services as a suite or on a component basis, tailored to their needs.

“It’s unifying a lot of what we’ve built over the years,” said Mayadas. “Databricks is the engine that gives us the ability to scale and leverage a bunch of complex technologies that help us surface data in AI in firms that want to drive agentic workflows.”

Corporate Actions

Intelligence Fabric for Capital Markets has been developed over a number of years and repeated testing has illustrated its utility, said Mayadas. He pointed to the success of a demo that showed how the concept manages corporate actions within workflows. This is usually a very complex process of multiple moving parts that demands real-time data responses. Much of this is often carried out manually and the identification of best models and data sets to solve the challenge is a common pain point.

“We had to figure out how should we model this corporate action? What’s the right data vendor? How do we link that corporate action into every single downstream source, a price master, an entity master? And then what’s the impact of all of that in terms of what our portfolio managers see, right?” Mayadas explained.

Rimes’ new fabric was able to streamline the process with minimal human intervention, “just human insight”.

Early Adoption

Mayadas said he sees validation of the company’s new approach in the fabric’s adoption by a large sovereign wealth fund, which is now helping the company to guide the evolution of the service.

One early expression of that will be an increase in provisioning for private assets investors, harnessing the fabric to bring the attributes and characteristics of those markets’ data in line with established data sets.

“Now that we’ve unified these components together on Databricks and we’re using AI to really increase the rate at which we can build new capabilities on that, we can now master complex private asset data on the same platform and bring a lot of the discipline that we see in the public markets to private markets data,” said Mayadas.

“We can bring a lot of the public discipline in the public asset world to the private markets world on a unified Databricks platform and ultimately surface a total portfolio view for the customer.”