IT Tech Pulse Exclusive Interview with Michael Jack Chief Revenue Officer and Co-Founder of Datadobi

Michael Jack, Chief Revenue Officer and Co-Founder of Datadobi, discusses how unstructured data visibility, governance, and AI readiness are becoming critical to enterprise security, cost control, and digital transformation.

Michael, you’ve spent 22-plus years in data storage, including time at EMC building Centera, before co-founding Datadobi in 2010. What convinced you unstructured data specifically, not storage capacity broadly, was the real problem?

At EMC, we were solving capacity and retention problems, how much you could store, how cheaply, how compliantly. What became obvious over time is that capacity was never the hard part. The hard part was that nobody actually understood what data was actually being stored. You can build infinite cheap storage and still have no idea what you’re storing, why, or whether it matters. That’s a fundamentally different problem than capacity, and it’s the one we set out to solve with Datadobi: not where data lives, but whether an organization actually understands what data exists well enough to make good decisions about it.

You’ve said orphaned data presents a serious security threat because it can contain sensitive information exploited by unauthorized individuals. Can you walk us through how a routine storage hygiene problem becomes a genuine board-level liability?

Orphaned data is data nobody’s accountable for anymore, the owner left, the project ended, but the files remained. Nobody’s watching who accesses it or what’s in it. That’s exactly the kind of data that ends up in a breach report, because it’s unmonitored and often forgotten. It only becomes a board-level issue once something happens, an audit, an incident, a regulator asking questions, and the organization realizes it can’t answer even basic questions about its own data. At that point it’s no longer a hygiene problem, it’s a governance and risk failure, and boards take notice of those.

Most IT leaders treat storage as a hygiene issue owned by infrastructure teams. What’s the actual trigger event a breach, an audit, or an AI project that forces it onto a CRO’s or board’s agenda?

Historically it’s been breaches and audits, something goes wrong and suddenly everyone wants to know what data existed and who could access it. What’s changed is that AI has become its own trigger, often a faster one. The moment an organization thinks about feeding an AI data lake with enterprise data, they discover there are important questions that can’t answer: what data is this, is it accurate, is it safe to use. That gap shows up immediately and visibly, which is why AI initiatives are pulling data management onto the board agenda faster than compliance ever did.

Organizations are discovering they can’t feed AI models reliably without knowing what data they have or whether it’s safe to use. What’s the single most common AI-readiness assumption CIOs get wrong right now?

The assumption that having data is the same as being ready to use that data. Most CIOs can’t even say how much unstructured data they have. Very few can tell you which of it is accurate, current, duplicated, sensitive, or actually relevant to the use case they’re building. That distinction, between having data and understanding it, is exactly where AI projects stall. You can’t align data to a downstream AI pipeline if you don’t understand what you’re aligning in the first place.

You’ve said AI will revolutionize data management and enhance data quality and accuracy. Does that same AI need clean, governed unstructured data to work at all, or can it help clean up the mess itself?

Both things are true, and that’s the tension organizations are living in right now. AI absolutely can help understand and classify data at a scale but humans can’t. But AI models built on ungoverned, unreliable data just produce ungoverned, unreliable outputs faster. So AI is simultaneously the reason organizations need to get their data house in order, and one of the tools that helps them do it. The operating discipline still has to come first, understanding what you have before you trust AI to act on it because once AI has access to it, that is a very hard thing to undo.

Make the economics case plainly: what does it actually cost an enterprise, in dollars, risk, or wasted AI spend, to keep ignoring unstructured data versus investing now in getting ahead of it?

The cost shows up in three places. You’re paying to store data you don’t need on infrastructure that’s more expensive than it should be. You’re carrying risk on data nobody’s tracking, which gets expensive fast the moment there’s an incident. And increasingly, you’re wasting AI investment, feeding models data that’s redundant, stale, or simply wrong, then paying again to fix the output or rebuild the project. Getting ahead of it isn’t a cost center, it’s what makes the storage, governance, and AI spend you’re already making actually pay off.

IT teams pay for storage they don’t control, since data owners decide what’s kept rather than IT. How much of the cost of doing nothing comes from that mismatch between who spends and who decides?

Honestly, a lot of it. IT carries the bill, but the decisions about what gets kept, moved, or deleted usually sit with business owners who aren’t thinking about the storage tab. That disconnect is exactly why treating this as a series of one-off infrastructure projects doesn’t work, it leaves the two sides talking past each other. What has to happen is a shared framework where business context and IT execution are actually connected, so decisions about data reflect both what it’s worth and what it costs to keep.

You said enterprises need data management capabilities that neither storage vendors nor hyperscalers fully address, and that this market won’t build itself. What’s the gap you’re describing, and what should organizations call it?

The gap is that storage vendors solve infrastructure problems and hyperscalers solve compute and platform problems, but neither is built to answer the question organizations actually have now: what does my data mean, and what should happen to it. That requires a discipline that sits above infrastructure: understanding data, aligning its treatment with business intent, then executing that at scale, continuously, not as a one-off project. Increasingly, the sharpest version of that question is being asked in service of AI: can this data be trusted and is it ready to use. That’s the capability the market still needs to build.

Thank you, Michael, for taking the time to share your insights with us.

Write to us [⁠wasim.a@demandmediaagency.com] to learn more about our exclusive editorial packages and programmes.