Fragmented data is stalling enterprise AI deployments before they ever ship

Enterprises are pushing past AI experiments and demanding production-grade deployments, but fragmented data and poorly scoped agents are stalling progress. The question is no longer whether AI agents can execute tasks — it’s whether the underlying data fabric can give them enough context to execute the right ones.

That context problem sits at the center of how process-centric AI is being built at Appian Corp., which is betting that a governed AI data fabric — not raw model power — is what separates useful agents from unpredictable ones, according to Mark Talbot (pictured), director of AI architecture at Appian.

“With your AI architecture, the salient components are first the data — the context that you provide to the AI — and that all exists in our data fabric,” Talbot said. “With our data fabric, you can draw out relations between your support cases [or] between your knowledge base articles that you have for a support application. You also have your tools that are in place. Agents need tools to do work. Your tools are your existing actions and your existing processes.”

Talbot spoke with theCUBE’s Dave Vellante and Alison Kosik at Appian World 2026, during an exclusive broadcast on theCUBE, SiliconANGLE Media’s livestreaming studio. They discussed how Appian’s AI data fabric powers agentic AI, the boundaries between agents and workflow and real-world customer adoption patterns. (* Disclosure below.)

AI data fabric as the connective tissue for agentic workflows

Appian’s AI data fabric functions as more than a catalog — it is the application platform around which workflows, agents and human tasks are organized. Rather than migrating data, Appian synchronizes it in place, building a knowledge graph that maps relationships across systems such as Salesforce Inc. and SAP SE without requiring a lift-and-shift, Talbot explained.

“Procurement data lives in many different systems,” he said. “What we do with data fabric is we synchronize that data within Appian so you can work with that to perform your semantic search — to perform your AI-powered decisions with our agent technology.”

That approach is already showing up in the field. Working with a wealth management firm in Australia, Appian’s team conducted a task audit to identify which human-driven IT support steps could be handed to an AI agent. The agent was given a goal, a set of instructions and access to the AI data fabric, then handled first-level triage autonomously — searching prior support tickets and knowledge base articles without human involvement. That pool of enterprise context is only growing: Appian’s recently announced partnership with Snowflake Inc. extends the data fabric’s reach further across enterprise systems. But wider context doesn’t mean looser control — where errors are costly or compliance is involved, deterministic workflows stay in charge.

“These multi-agent platforms need to be governed by workflows, especially when the cost of error is high,” Talbot said. “You need an overarching workflow process that delegates work to the individual agents. For example, you might have an individual agent that reads a report from a physician, interprets if it’s high risk or low risk, and then a human still in the loop is part of your overarching process to review that.”

Here’s the complete video interview, part of SiliconANGLE’s and theCUBE’s coverage of Appian World 2026:

(* Disclosure: TheCUBE is a paid media partner for Appian World. Neither Appian, the sponsor of theCUBE’s event coverage, nor other sponsors have editorial control over content on theCUBE or SiliconANGLE.)

Photo: SiliconANGLE