Ereignis

2026 Gartner Data & Analytics Summit: How to Successfully Move AI from Pilot to Production

Michael Curry, President of Data Modernization at Rocket Software, shared the 5 key shifts needed to scale AI from experiment to enterprise impact. Explore the key insights from his session below.

Take your session insights into your organization

 

Rocket Booth

 

At the 2026 Gartner Data & Analytics Summit, we heard that enterprises don’t have an “AI problem”—they have a data access, trust, and control problem. Pilots stall when they hit legacy data, protected documents, and governance constraints. We heard that how we respond to these problems ties directly to big challenges like scaling AI, managing disruption, and balancing risk with governance. Rocket is here to help you turn those insights into action—unlocking your data’s potential to drive real, lasting success.

 

Three actions you can execute quickly

Organizations can make real progress fast by:

  • Defining trust SLAs for their 10 most critical datasets.
  • Implementing governed retrieval patterns across both data and content.
  • Standing up early semantics + data product models that AI agents can use immediately.

These create momentum, reduce rework, and shrink AI experimentation risk. Rocket is here to help you turn those insights into action—unlocking your data’s potential to drive real, lasting success.

Sprechen Sie mit einem Experten

Key takeaways from the 2026 Gartner Data & Analytics summit

  • AI readiness starts with trust, not models: Trust in data quality, lineage, security, and context. Organizations must formalize trust SLAs that guarantee consistent, governed data delivery across all modernization and AI initiatives.
  • Governed retrieval needs to be a strategic capability: AI forces enterprises to manage retrieval across structured data, unstructured content, and mainframe‑resident knowledge. The winners will be those who treat governed retrieval as a first‑class architectural pillar, not a bolt‑on control.
  • Data Sovereignty Must Shift From “Location‑Based” to “Use‑Based”: Traditional sovereignty frameworks break under AI’s dynamic access patterns. Enterprises need adaptive, use‑based sovereignty, where rules follow the context of use — not the system, region, or storage boundary.
  • Semantics Are the New Interface for AI Agents: AI agents can’t thrive on raw tables or legacy schemas. They need agent‑ready semantics — clear meaning, relationships, and intent — expressed through governed metadata, knowledge models, and well‑designed data products.
  • Data Products Must Be AI‑Aligned, Not Just Analytics‑Ready: Enterprises need data products enriched with machine‑readable semantics, built for retrieval‑augmented generation (RAG), agent orchestration, and continuous ML intelligence — not just dashboards or ETL outputs.
Verwandte Inhalte
Video

The AI Advantage: Maximizing Mainframe Data for Analytics

Trotz Einschränkungen lassen sich durch Mainframe-Daten Wettbewerbsvorteile erzielen.

Jetzt ansehen
Insight

Von Datensilos zu Strategien: Das wahre Potenzial von Daten freisetzen

Die Rolle des Datenmanagements besteht heute nicht mehr nur darin, Daten zu sammeln und zu speichern, sondern darin, sie zu sehen, ihnen zu vertrauen und maximalen Nutzen für geschäftliche Vorteile daraus zu ziehen, unabhängig davon, wo sie gespeichert sind.

Jetzt lesen
Insight

Beyond the Buy-Side: Your Mainframe Data Isn't Stuck. Your Strategy Is

Break past data bottlenecks and turn core systems into advantage.

Jetzt lesen