A new startup called June emerged from stealth today with $20 million in pre-seed funding to tackle what may be the most persistent and underreported problem in enterprise AI: companies buy the tools, then nothing works.

The round was led by Marc Benioff’s Time Ventures, with backing from Michael Dell, Box CEO Aaron Levie, and CrowdStrike CEO George Kurtz. The founding team includes Efrat Rapoport, a former Salesforce executive, along with Ohad Hen, Barak Goldstein, and Idan Tsitiat.

The investor lineup reads like a who’s who of enterprise software. That’s not accidental. Every one of them has watched the same movie play out across thousands of companies: executives announce an AI strategy, sign contracts with OpenAI or Salesforce AI or ServiceNow, and then spend months watching their teams struggle to make any of it actually function inside their existing systems.

The Problem June Is Solving

The AI deployment gap is real, and it’s costing businesses more than just time.

Gartner estimates that only 17 percent of organizations have fully deployed AI agents, even though the vast majority have purchased AI tools or started pilots. The disconnect isn’t about the quality of the models. It’s about the messy reality of enterprise software: legacy systems, data in seventeen different formats, approval workflows that predate smartphones, and integration requirements that no AI vendor built for.

Rapoport framed it plainly: “AI paradoxically increases the demand for professional services.” The smarter the tools get, the more complex the implementation becomes, because now you’re trying to wire autonomous systems into infrastructure that was designed for humans to operate manually.

June’s approach is to automate the implementation work itself. Rather than hiring armies of forward-deployed engineers to manually integrate AI into enterprise stacks, June builds agents that identify bottlenecks in deployment workflows and resolve them programmatically. The target systems are the ones that appear in almost every enterprise: Salesforce, ServiceNow, Workday, and the dozens of tools that orbit them.

Why This Matters More Than the Model Race

Most of the attention in enterprise AI has focused on which model is most capable. But capability is table stakes. The real question for most businesses isn’t “which AI is smarter” but “how do I make any of this work inside my company.”

That’s why the investor interest here is telling. Benioff built Salesforce on the insight that software deployment itself was broken, and that cloud delivery could change the economics of implementation. Dell has seen the same pattern from the hardware and infrastructure side. Levie and Kurtz have each watched enterprise customers struggle to operationalize powerful tools inside complex environments.

The bet with June is that the implementation problem is now big enough, and AI capable enough, to automate the implementation layer itself. In other words, deploy agents to help deploy AI.

This mirrors a broader trend. The current wave of enterprise AI investment is shifting from “build better models” toward “make existing investments actually work.” That’s where the real operational value lives, and it’s where companies like June are positioning.

What This Means for Business

If you’ve bought AI tools and can’t get them running at full potential inside your business, you are not alone. The most common failure mode in enterprise AI right now isn’t bad technology. It’s the gap between a vendor’s demo environment and your actual systems.

A few practical takeaways from this development:

The implementation gap is getting venture attention. When Marc Benioff and Michael Dell write checks into a deployment-focused startup, it signals that the bottleneck is real and widely felt. If two of the most experienced enterprise software investors in the world see it as a multi-billion-dollar problem, it’s worth taking seriously.

Agents will deploy agents. The idea that AI can automate its own implementation is no longer theoretical. It’s now the thesis of a well-funded company with a serious founding team. Expect this category to grow.

Custom implementation work is expensive. If your AI deployment has stalled because you’re waiting on a systems integrator or internal IT resources, that waiting has a cost. Every week of delayed deployment is a week of productivity sitting unrealized.

The systems you already have matter more than the next model. Businesses chasing the latest AI release while struggling to integrate what they’ve already bought are solving the wrong problem. The leverage is in making your current investment work, not adding to it.

The EDNA Perspective

This is exactly why operational advisory and hands-on implementation support matters. The data shows that AI tools alone don’t produce results. The outcomes come from how those tools are integrated into the workflows, data systems, and decision processes of a specific business.

Enterprise DNA’s Omni services exist precisely in this gap. Omni Ops builds AI agent workforces grounded in a business’s actual context, not generic demos. Omni Advisory helps leadership teams understand where AI can create real leverage in their specific operations before committing to deployments that may stall.

The fact that investors are now writing $20M checks to solve this problem validates what businesses are actually experiencing day to day: the deployment problem is real, it’s expensive, and it’s not going to solve itself.

June’s emergence is worth watching. But more importantly, it’s a signal that the era of “AI pilot” culture is ending. The question for every business leader is no longer whether to adopt AI but whether they have the right support structure to make it work.

If the answer is no, that’s the thing worth fixing first.