At the Gartner Data & Analytics Summit in Orlando, analyst Masud Miraz presented a 10-step framework for building a data fabric architecture - a layer he described as essential before enterprise AI use cases can produce reliable, contextual results. A 2024 Gartner survey of 247 data management leaders underscores the stakes: 77% put AI-ready data at the top of their investment priority list for the next two to three years, but far fewer are funding the architectural changes needed to deliver it.
Only 37% of those surveyed are upgrading their data management architecture, 40% are investing in active metadata tools, and just 25% are pursuing lakehouse initiatives. Miraz positioned data fabric as a long-term orchestration engine and control plane that spans analytical and operational stores, providing the missing link between raw data and AI workloads.
The data fabric blueprint
Miraz laid out three operational goals for the fabric. Flexible pipelines must detect schema and code drift without manual intervention. Augmented data engineering should automate repetitive work - "80% of current data engineering tasks are repetitive," Miraz said - freeing teams for higher-value work. Workload management, meanwhile, balances cloud costs against performance demands.
The fabric runs on three key inputs: metadata pulled directly from systems, active metadata insights that surface real-time intelligence, and knowledge graphs that add semantic understanding. Together, these components give AI models the clean, governed context they need to perform dependably.
A recurring investment gap
Editorial analysis from the summit noted that the disconnect between survey intentions and actual spending mirrors a familiar pattern in enterprise analytics. Organizations often chase surface-level AI use cases while deferring deeper investments in metadata, governance, and unified control planes. Many production machine learning systems remain fragile as a result, unable to reproduce the consistent results that business leaders expect.
What to watch
Industry observers should monitor adoption rates for active metadata tools and announced lakehouse initiatives, as well as vendor roadmaps that integrate knowledge graphs with metadata layers. Case studies that quantify reductions in schema-drift incidents or cloud cost improvements after deploying fabric-like control planes will offer the clearest signal of real progress. Follow-up surveys repeating Gartner's questions over the next year will reveal whether AI-data priorities finally translate into architectural spending.
Why this matters for management
For management teams, the data fabric discussion is not a technical side note - it is a budgeting reality. The survey shows near-universal recognition that AI-ready data matters, yet most organizations are not funding the control planes and metadata layers that make it possible. Shifting investment from surface AI features to foundational data architecture now will determine which companies can deploy models that are reliable, cost-efficient, and truly contextual at scale.
Your membership also unlocks:
