IBM’s Edward Calvesbert on AI-Ready Data in 2026

In the past year, most enterprises have been focusing on generative AI pilots. However, many are stuck at this stage due to inadequate data readiness. This article explores these challenges and insights from Edward Calvesbert, Vice President of Product Management for watsonx.data at IBM.

The Core Issues in AI-Ready Data

According to Calvesbert, the main issues are data quality and fragmentation. Enterprises often struggle with data trapped across silos that lack structure, metadata, and proper governance.

Predicted Shifts Reshaping Data Management

  • Hybrid cloud becoming the standard design pattern for enterprise scalability
  • Zero copy integration reducing duplication costs and improving time-to-value
  • Frontier models enabling easier combination of structured and unstructured data for new insights

The Market is Consolidating

Data platforms are converging around fewer vendors, but these vendors will build on open standards instead of closed ecosystems. IBM’s watsonx.data is positioned as a leading open hybrid-cloud data platform.

Challenges and Solutions for AI-Ready Data

Enterprises often lack unified access to structured and unstructured data, consistent governance, and a clear path from pilots to production with proper security, compliance, and cost-effectiveness.

The Stumbling Blocks for Gen AI Projects

  • Data fragmentation hinders information access and combination across sources and formats
  • Missing enterprise readiness features like security, compliance, and governance create deployment barriers
  • Consistent accuracy and reliability are crucial as enterprises move from informational use cases to analytics and agentic automation

Zero Copy Access: A Solution in the Making

Zero copy means querying data where it resides without moving or duplicating. It solves several critical problems by eliminating duplication costs, reducing ETL processes, and avoiding vendor lock-in.

The Future of Hybrid Cloud

Hybrid cloud is no longer a transitional state. Real-time latency requirements, compliance mandates, cost pressures, and concerns about hyperscaler lock-in are driving hybrid-by-design strategies.

Cost Efficiency and Portability

The growing frustration around analytics and AI costs is pushing the market towards greater portability. More workloads will be placed across different engines from various vendors to achieve optimal price-performance.

Watsonx.data Solutions

  • Multifunctional data engines like Presto, Spark, OpenSearch, and Cassandra
  • Native C++ and Jvector capabilities for optimal price-performance
  • GPU-accelerated execution to process unstructured and AI-generated data more efficiently

The Competitive Landscape

Data platforms are consolidating, but they will support open standards at multiple levels in the stack. IBM’s watsonx.data is positioned as a leading platform for enterprises.

Uncomfortable Truths for Data Leaders

  • Data estates remain too complex and fragmented to support AI at scale without unified access and governance
  • Agentic development and analytical tools require fluency among users with diverse skills