Robin Sutara, Field Chief Data Strategy Officer at Databricks, speaks with Robert Lutton, VP, Sales and Marketing, Sandhill Consultants, in a video interview about the organization, its Data Intelligence platform, how the platform aids data readiness, the importance of strategic partnerships, simplifying data infrastructure for AI acceleration, and how Databricks addresses the need for integrated governance.

Sutara begins by sharing an overview of Databricks, an 11-year-old organization founded by five PhDs from UC Berkeley, who were also the creators of Spark. It has evolved significantly over the years, and today, its main focus is helping organizations eliminate data silos across both structured and unstructured data.

This enables organizations to use the same datasets for both AI and business intelligence without needing to duplicate or transfer data.

Databricks has driven innovation through both strategic acquisitions and internal breakthroughs, all aimed at building an intelligent data platform. This “Data Intelligence Platform” enables organizations to not only build their AI solutions but also leverage embedded AI within the platform to reduce costs, improve operational efficiency, and unlock broader business value.

Laying the groundwork for AI: Data readiness, platform unification, and strategic partnerships

Moving forward, Sutara emphasizes the centrality of AI in today’s business discourse and the challenges organizations face in turning AI ambition into execution. She acknowledges how omnipresent AI discussions have become but also points out a persistent gap between strategy and action.

Many businesses, Sutara observes, are still grappling with fundamental questions of how to activate AI strategies and unlock true data value. For most organizations, this journey starts with getting their data in order. 

Sutara explains that without strong data foundations, companies cannot take full advantage of recent technological advances in AI. This is where she sees Databricks’ Data Intelligence Platform making a significant impact.

According to Sutara, it helps democratize access to data and AI by unifying disparate systems that include structured data in data warehouses, unstructured data in data lakes, isolated AI tools, and disconnected governance systems.

Next, Sutara highlights the growing trend toward complex agentic AI systems. As companies mature in their AI adoption, they need platforms that support more sophisticated, integrated AI solutions, she adds.

Further, Sutara poses a critical question for organizations: “Can your platform support both enterprise-wide models and the smaller, proprietary agentic models unique to your business?” She stresses the importance of being able to use all types of data — structured and unstructured — within a unified system.

Partnerships are key to enabling this vision, says Sutara. Referring to a recent collaboration, she says, “Our recent announcement with the partnership with SAP, for example, I don’t think there’s any customer that I’ve worked with that doesn’t have some level of SAP within their environment.”

Sutara notes that SAP’s data is highly organization-specific and valuable. With Databricks, companies can now unlock that value and integrate it deeply into both their AI and BI initiatives.

Thereafter, Sutara reflects on Databricks’ evolving role: “How do we continue to disrupt ourselves in order to help organizations realize the full value of their data and prepare for the future?” That, she says, is the question that drives the platform strategy forward.

Simplifying data infrastructure for AI

Next, Sutara sheds light on the real-world complexity most organizations face when managing their data landscapes and how Databricks is evolving to meet those challenges. She describes Databricks’ progression as a mission to continuously remove the barriers between disparate data systems and make it easier for organizations to manage, access, and act on their data.

“For us, it’s always been about how do we help you eliminate those data silos? How do we help you break down the barriers that exist between those?” she says. “Because we know it’s currently a fragmented system, and for most organizations, it’s not even just like one warehouse. It’s multiple warehouses. It’s not just one lake; it’s multiple lakes.”

Sutara then points out that corporate events like mergers, acquisitions, and divestitures only intensify that fragmentation. Each change introduces new layers of complexity to data management, making it even harder for companies to gain a unified view of their operations.

At the core of Databricks’ platform strategy is the creation of a single source of truth — a unified layer that spans across warehouses, lakes, and various data systems within the enterprise.

“If you can break down those silos, you’re going to accelerate your AI. You’re going to accelerate that value out of the data and the insights that you’re looking to derive, and you’re going to reduce costs.”

The tangible benefits of simplifying data platforms include faster time to insight, quicker innovation cycles, accelerated time to value, and improved cost efficiency across the organization.

Growing need for integrated AI governance

Speaking of governance, Sutara states that Databricks considers governance to be fundamental for any platform and that it cannot be a separate ecosystem. The organization’s data intelligence platform has a Unity Catalog, which is its governance foundational layer and covers the entire platform, she adds.

Further, Sutara mentions that most organizations face uncertainty around legislative and regulatory requirements when addressing AI governance. She references the EU AI Act as a clear example of formal regulation taking shape while noting that in the U.S., the national approach remains unclear. “But lots of U.S. states are starting to pass their own legislation that companies need to comply with,” she adds.

Organizations, Sutara maintains, are now trying to navigate key governance questions:

• How to trace end-to-end lineage?
• What datasets were used, and what models did they feed into?
• What were the weightings and algorithms behind those models?
• What data products and services are being powered by those models — and what do they feed into next?

These concerns make transparency and explainability essential. “Your governance layer has to be able to handle all of those aspects,” she says. “You need to be able to identify not just data, but also the AI that the data is being used on.”

Concluding, Sutara states that for Databricks, that means it has to be integrated into the platform. The governance layer must be able to support the organizations and meet the expectations and maintain the trust of the people about how their data is used and what it is feeding into.

CDO Magazine appreciates Robin Sutara for sharing her insights with our global community.