Eric Kalin Highlights Why AI Readiness Starts With a Strong Data Foundation

Alpharetta, Georgia, November 9th, 2026, ZEX PR WIRE,

As businesses move from AI experiments to systems that can take action, global technology executive Eric Kalin says the data beneath them deserves more attention.

Businesses are moving quickly from experimenting with generative artificial intelligence to exploring agentic systems that can participate in workflows and take action. Eric Kalin believes that shift is making an older enterprise challenge newly important: understanding the data those systems rely on.

After years of cloud adoption, application growth, system changes, and business expansion, enterprise information rarely sits neatly in one place. It can be spread across departments, databases, older systems, and multiple cloud environments. Some information may be current and well managed. Other data may be duplicated, outdated, difficult to access, or subject to different security requirements.

For Kalin, that means the conversation about AI readiness increasingly needs to become a conversation about data readiness.

“There is a tendency to start with what the AI can do,” Kalin said. “I think you also have to ask what you’re giving it to work with. Where did that information come from? Can you trust it? Is the system supposed to have access to it? Those questions become even more important when AI moves from generating an answer to taking an action.” Kalin has more than 25 years of experience across enterprise technology, cloud computing, data, consulting, sales, and business transformation. His career includes leadership roles at IBM, Accenture, Amazon Web Services, Microsoft, and Oracle. That experience has allowed him to see how enterprise technology environments have become more powerful while also becoming more complex.

Having More Data Does Not Mean Being Ready for AI

Organizations today can collect and retain enormous amounts of information. Kalin argues that volume alone says little about whether that information can support an AI application.

“Having more data doesn’t make you more ready for AI,” Kalin said. “Understanding your data does.” Before scaling an AI initiative, organizations may need to determine where relevant information is stored, how current it is, whether different systems agree with one another, and who is responsible for maintaining it.

The challenge becomes particularly important when information about the same customer, product, transaction, or business process exists in several places.

A sales team, for example, may work from one set of customer information while finance and operations rely on different systems. Each system may serve its purpose well, but an AI application attempting to work across the organization may encounter conflicting or incomplete information.

Kalin believes these situations illustrate why AI cannot be separated from the wider technology environment in which it operates.

Multicloud Environments Add Another Layer

The growth of cloud computing has given businesses greater flexibility in how they build and operate technology. It has also contributed to environments where applications and data may be distributed across different platforms.

For organizations operating across multiple clouds, the challenge is not necessarily to bring every piece of information into one location. It is to understand how information moves, which systems should be able to use it, and what controls need to follow it.

Kalin’s career has included extensive experience with cloud technology and global enterprise customers. He believes the same practical thinking that became important as cloud adoption matured is now becoming relevant to AI.

“Eventually, every major technology conversation gets more practical,” Kalin said. “You move from asking what is possible to asking how it actually works inside your organization. With AI, data is a big part of that second question.” That includes understanding which information matters for a particular AI application, rather than assuming every available source needs to be connected.

Agentic AI Raises the Stakes

Data quality and access already matter when generative AI is being used to summarize information, answer questions, or assist employees. Agentic AI adds another consideration because systems may increasingly participate in workflows and take action.

“With generative AI, you might be concerned about an inaccurate answer,” Kalin said. “With an agent participating in a workflow, you also have to think about what happens after that answer. What can the system access? What is it allowed to change? When does a person need to approve an action? The data and governance questions become very practical very quickly.” This is where security, permissions, and ownership become closely connected to data quality.

Making information available to AI does not mean making everything available without restriction. Organizations still need to determine who owns sensitive information, which systems can access it, and where human review remains necessary.

Kalin believes those decisions should involve more than technology teams alone. Business leaders often understand how information is actually used. Security teams understand the risks surrounding access. Data teams understand quality and structure. Bringing those perspectives together can help uncover problems before an AI application is widely deployed.

Perfect Data Is Not the Goal

Kalin also cautions against interpreting data readiness as a requirement to solve every information problem before experimenting with AI.

Few large organizations have perfectly organized data. Waiting for perfection can create its own barriers to progress.

Instead, early AI projects can help businesses identify which data issues actually matter for a specific use case.

“You don’t need to fix every piece of data in the company before you start,” Kalin said. “You need to understand the data that matters for what you’re trying to accomplish. Start there, learn from it, and build from what you find.” That approach keeps the focus on a defined business need rather than attempting a massive data project with no clear connection to an outcome.

As companies move from AI experimentation toward broader use of generative and agentic systems, Kalin expects questions about data quality, access, security, and governance to become harder to separate from the AI conversation itself.

The organizations with the largest amounts of information will not automatically be the ones best positioned to use AI effectively. Understanding what that information means, where it lives, how reliable it is, and how to use it responsibly may prove just as important as the AI technology placed on top of it.

Eric Kalin is a global business leader and technology executive with more than 25 years of experience across enterprise technology, cloud computing, data, artificial intelligence, consulting, sales, and business transformation. His career includes leadership roles at Oracle, Microsoft, Amazon Web Services, Accenture, and IBM. Based in Alpharetta, Georgia, Kalin has worked with global organizations and cross-functional teams throughout major periods of change in enterprise technology. His interests include data platforms, generative and agentic AI, multicloud environments, business strategy, and the connection between emerging technology and practical business needs.