The Future of Data Intelligence: Insights from revAlation to You

The enterprise data landscape is undergoing a seismic shift. As artificial intelligence transforms how organizations consume and produce information, data teams face mounting pressure to deliver measurable business value rather than simply cataloging assets. 

At Alation's recent revAlation webinar, CEO Satyen Sangani and 84.51°’s VP of Architecture Nate Sylvester shared insights on navigating this transformation—and unveiled powerful new capabilities that position data products as the foundation for successful AI initiatives.

The evolution from data for data's sake to business outcomes

For the past decade, data teams operated in an environment where "data was seen to be good in and of itself," Sangani explained. Zero interest rates and expanding IT budgets meant data professionals could focus on infrastructure without constant pressure to demonstrate ROI.

That era has ended. Today's data leaders must connect their work directly to business outcomes—whether regulatory compliance, operational efficiency, or revenue growth. This shift demands a fundamental reimagining of data management strategy.

Data products: The missing link between data and value

Enter data products: datasets designed to drive value. As Sangani described it, "a data product is simply a dataset with a purpose."

Swire Coca-Cola provides a compelling example. The beverage distributor struggled with dozens of reports measuring OTIF (on-time, in-full delivery)—each with slight calculation variations. Nobody knew which metric to trust. By implementing a data product approach, they created a single source of truth for OTIF, then worked backwards to implement the governance, data quality checks, and lineage required to support that outcome.

Nate Sylvester's team at 84.51° (Kroger’s data science subsidiary) has taken this concept even further, developing a three-tiered data product architecture that balances reusability with business focus.

Kroger's data product architecture: A practical framework

Sylvester's team evolved their data strategy over several years, moving from centralized data warehousing to full federation and finally to a product-centric approach. Their current framework includes three distinct layers:

• Source data products serve as the foundation, capturing core business entities like POS transactions, inventory data, and supplier interactions. These change infrequently and provide the raw material for downstream use cases.

• Edge analytics products address specific business questions like demand forecasting. These iterate rapidly and focus on narrow use cases, feeding directly into dashboards and decision-making processes.

• Intelligence products aggregate source data into broader business concepts.

Kroger's “customer sale” data product exemplifies this final layer. It stitches together data from multiple systems—point-of-sale transactions, customer profiles, item attributes, and even shopping modality (e.g., in-store or online)—to create a unified, contextual understanding of each customer purchase.

As Sylvester explained:

“A customer sale is a series of these source data products stitched together. We have POS data, customer data, item data, maybe even location and modality. That customer sale data product introduces a new concept—a new construct—into the organization around what it means to make a customer sale. From there, we can build insights on top of that. We can start to understand customer behavior, forecast demand, and feed a lot of downstream analytics because these concepts are now related.”

By integrating these layers, Kroger has created a framework where context becomes the differentiator. The “customer sale” data product allows analysts and AI systems alike to uncover deeper patterns in shopper behavior—how customers interact across channels, which promotions drive repeat visits, and where loyalty programs make the greatest impact. This unified view transforms raw data into actionable intelligence.

This architecture enables Kroger to balance stability with agility—foundational products provide reliable building blocks while edge products evolve quickly to meet changing business needs.

Introducing Chat with Your Data: AI meets data intelligence

One of the most significant announcements from revAlation was Alation's new Chat with Your Data feature, part of the company's broader Data Products Marketplace. This capability allows users to ask natural language questions and receive answers grounded in governed, contextualized data.

The feature represents a fundamental shift in how people interact with data. "Instead of going to sites like Wikipedia or frankly even Google, people are instead going to ChatGPT," Sangani noted. The same transformation is happening in enterprise data management. Rather than navigating through a catalog interface—or struggling with disparate or duplicate data sources— users can now chat directly with their enterprise data, just as an individual might interact with ChatGPT, but within the governed environment of the Alation Data Catalog. The experience blends conversational convenience with the rigor of structured, compliant data.

Metadata delivers the context that makes this possible: defining business terms, clarifying metrics, and mapping which regulations apply to specific datasets. This ensures every AI-driven answer reflects not only the right data, but the right meaning.

Four pillars of AI-ready data management

Alation's product roadmap centers on four strategic pillars designed to help enterprises succeed with AI:

1. Declarative governance

Rather than procedural governance committees where "people don't show up" because they don't understand the value, declarative governance lets teams specify business outcomes and work backwards to necessary policies. Forthcoming updates will exemplify this approach, using AI to suggest appropriate checks and identify critical data assets.

2. AI-ready data products

Data products serve as containers that make data consumable to AI systems. They provide the context, governance, and structure that models need to operate effectively. Alation's new Data Product Builder helps teams define product purpose, identify required attributes, select appropriate datasets, and implement necessary governance—all through a guided interface.

3. Analytical agents

Alation's Agent Builder enables teams to create metadata-aware AI agents that automate analysis and decision-making. The platform provides out-of-the-box (OOTB) agents for chat, querying, dashboard creation, and curation — as well as the ability to build no-code custom agents with built-in evaluations, infrastructure, and vendor-neutral flexibility. For teams seeking deeper control and extensibility, the Agent SDK offers a programmatic experience designed for high-touch customization and integration.

JLL, the commercial real estate firm, demonstrates the power of this approach. They rebuilt their lease renewal process around AI agents fed by data products, achieving dramatic results: "4x faster" decision-making with customers, "18% fewer vacancies," and "70% less manual work" for real estate agents.

4. Open and extensible knowledge layer

Alation’s Agentic Knowledge Layer is a unified, metadata-driven foundation that enables AI agents to access, understand, and act on enterprise data with precision and trust.

Unlike traditional catalogs or BI tools, the Agentic Knowledge Layer powers autonomous and semi-autonomous AI agents by combining data governance, metadata, and data products into a declarative, automated system.

This layer bridges a crucial gap: while LLMs excel with unstructured text, they often lack awareness of enterprise schemas, joins, and relationships—leading to hallucinations and unreliable SQL. Grounding AI in metadata through the Knowledge Layer ensures accuracy, compliance, and explainability.

In short, without an Agentic Knowledge Layer, enterprise AI sounds right but often isn’t. With it, AI becomes reliable, explainable, and aligned to business context.

The layer integrates the aforementioned pillars:

• Metadata, which provides semantic context and lifts Text2SQL accuracy by up to 30%.

• AI agents, governed and customizable for enterprise use.

• Data products, curated for reuse and quality.

• Declarative governance, embedding compliance into the fabric of automation.

Together, these capabilities power Chat with Your Data, Agent Builder, and other Alation features—delivering accuracy, adaptability, and data sovereignty across 100+ connected systems.

The company’s new “zero data” option further enhances security by keeping all analytical data within customer firewalls, communicating directly with browsers without passing through Alation’s cloud infrastructure.

Getting started: Practical advice from the experts

When asked how to win leadership support for data products and governance, Sylvester emphasized starting small and tying initiatives to business outcomes. "You can't have good AI without good data," he noted, suggesting data leaders attach their programs to executive enthusiasm for AI initiatives.

The key is trusted AI—and that trust comes from metadata. Metadata provides the definitions, lineage, and business context that make AI-generated outputs accurate and meaningful.

Both speakers warned against "boiling the ocean"—attempting to organize all data before demonstrating value. "The worst thing you can do is start talking about data capabilities and features instead of just talking about business outcomes and business value," Sangani advised. "Talk about what they care about, which is the strategy they're trying to drive and the outcomes they're trying to drive, and then work backwards."

For skill sets, Sylvester described data product teams as cross-functional, requiring product managers, information architects (data analysts), and engineering talent. However, he noted that modern tools and AI assistants are reducing the programming knowledge required for pipeline development.

The path forward: Data as strategic enabler

As enterprises race to implement AI, the quality and accessibility of their data will determine success or failure. Alation's vision positions the data catalog not as a destination but as an intelligent layer that powers AI agents, ensures governance, and connects data directly to business outcomes.

"Data teams become a fundamental enabler of what the outcomes are that the business is trying to drive, and ROI becomes a straight line as opposed to something that people have to guess about," Sangani concluded.

For organizations ready to move beyond cataloging toward true data intelligence, the message is clear: start with business outcomes, build purpose-driven data products, and leverage AI to transform how data creates value. The tools are ready. The question is whether your data strategy is.

Watch the full webinar here.

Schedule a demo with us today.