As AI moves from experimentation to production, many organizations are discovering that their data foundations are not ready to support it.
Critical data is scattered across cloud platforms, operational systems, and data lakes, making it difficult to deliver trusted data for AI.
Without scalable integration and governance frameworks, inconsistent definitions, data quality issues, and limited visibility can undermine AI outcomes and limit the impact of AI initiatives.
DBTA recently held a webinar, AI-Ready Data: Building Scalable Integration and Governance Frameworks, with industry experts who explored how organizations are building AI-ready data foundations.
Jeff Hainsworth, senior product marketing manager at insightsoftware, pointed out the hype/result disparity by citing the following statistics:
- 88% of AI pilots never make it to production.
- 40% or more of AI projects could be scrapped by 2027.
- 95% of generative AI delivers no measurable ROI.
- 74% of companies haven’t seen value from AI.
- 25% say complex data is their biggest AI roadblock.
AI breaks when the data cannot be trusted, Hainsworth explained. Without data access, AI hallucinates. AI needs access to data instead of just copies. However, if there is no governance, there is no access. Furthermore, AI requires data prep.
Hainsworth introduced insightsoftware Simba Intelligence, where trusted data and ai finally connect. “Your AI is lying to you. We built the truth layer that stops hallucinations,” Hainsworth said.
Siloed data creates a risky context gap for AI, said Aaron Qayumi, senior product marketing manager at Reltio. Data silos and misaligned policies fragment context for data consumers. AI agents operate without context and guardrails, making poor decisions. AI initiatives stall or require hours of ongoing humanin-loop intervention.
Without data unification, AI lacks trusted core data, Qayumi noted. Missing unified context means AI agents are fast, confident, and wrong.
A real-time system of context solves this problem. It unifies entities, relationships, and meaning. Key requirements for a system of context include:
- Context-rich knowledge graph with entities, metadata, relationships
- Open, interoperable platform with MCP and API-first architecture
- Real-time, event-driven performance in milliseconds at scale
- No-code deployment of unified, agent-ready data products at scale
It should also be governed by design, offering integrations that inherit the platform’s security, RBAC permissions, lineage, and audit policies—for compliance and transparency across every flow.
Agents fail when data is poorly understood, said David Thain, product marketing lead at Informatica from Salesforce. The Informatica Intelligent Data Management Cloud offers a unified metadata foundation. Empower the agentic enterprise by embedding AI governance in every phase of the development lifecycle, Thain said.
After working with more than 12,000 of the world’s leading organizations, we understand the importance of Data Integrity for business success in the AI era, noted Carlos Arias, senior director, product management at Precisely.
AI is showing up everywhere, but its ROI is elusive, Arias explained. The opportunity is massive, but without trusted, unified, and governed data, AI can't deliver business impact. The winners will be those who prioritize their data strategy.
Integration and governance break down when data is:
Trapped: Hard to access and find data across hybrid IT landscape
Outdated: Backward-looking, periodic
Inconsistent: Multiple versions of data and formats that block the truth
Non-compliant: No traceability, unverified, uncontrolled autonomy
Expensive: Held back by specialized skills and manual processes
Arias offered the five principles that define AI-ready data, which include:
- Access Your Data: Connect all disparate data from across your IT landscape and make it fully discoverable.
- Operate in the Now: Always run in a known state; ensure data is continuously refreshed and up to date.
- Shape Data for Purpose: Deliver verifiable data that is relevant, complete, and with context for every AI workflow, action, decision.
- Elevate Governance: Put in guardrails to mitigate growing AI-related risks.
- Lower Cost Structure: Leverage AI to remove manual processes and specialized skills that cost the organization time and money.
For the full webinar, featuring a more in-depth discussion, Q&A, and more, you can view an archived version of the webinar here.