The AI Deployment Problem Nobody Talks About
Here's the thing about enterprise AI: building an agent template is easy. Getting it to actually work inside a Fortune 500 company's existing infrastructure? That's where everything falls apart.
According to Efrat Rapoport, a former Salesforce executive who just emerged from stealth with her new company June, "Before AI can create value, someone has to deal with legacy systems." She's not exaggerating. Enterprises have fragmented data scattered across Salesforce, ServiceNow, Databricks, Workday — you name it. Years of technical debt pile up. Duplicate database fields exist because different teams created them independently. Complex workflows choke on their own complexity.
The industry's answer to this mess? Hire more people. Specifically, forward-deployed engineers (FDEs) — specialists who literally drop into companies to get AI systems up and running. It's ironic, Rapoport notes, that "AI, paradoxically, increases the demand for professional services." The more AI tools companies buy, the more consultants they need to make them work.
Forward-Deployed Engineers: The Industry's Answer
Paul Akinmade, chief strategy officer at CMG, a major U.S. mortgage lender, knows this pain firsthand. Last year, he stood at Salesforce's annual conference and promised his company would return with 100 agents running on their platform. Fast forward to now, and he was hitting walls — meeting with architects, talking to forward-deployed engineers, consulting everybody he could think of — without making real progress.
"We spent weeks hitting a wall," Akinmade said. "We were stuck."
The problem isn't that AI models don't work. It's that they have to integrate with legacy systems that weren't designed for them. As Rapoport puts it: "How does an agent know how to operate when you have 10 duplicate database fields that say the same thing, and different teams are using them?"
Building an agent template is the easy part. The hard part is getting it to work with the mess underneath.
June's Automated Approach to Agent Integration
June emerged from stealth on August 3, 2026, raising $20 million in pre-seed funding led by Marc Benioff's Time Ventures. Additional backers include tech luminaries like Michael Dell, Aaron Levie, and George Kurtz. The company declined to share its valuation.
The four founders — Rapoport, Ohad Hen, Barak Goldstein, and Idan Tsitiat — previously started Bonobo AI, a voice-to-text service launched in 2017. Salesforce acquired them two years later, and the team worked on AI initiatives for several years before setting out on their own again after watching customers struggle to bring AI into their existing platforms.
Their potential was clear enough to their investors that "we didn't even have a deck for this raise," Rapoport says.
June's platform works differently than most AI tools. Instead of asking companies to hire FDEs or consultants, it scans their existing systems to understand business processes, finds bottlenecks, and then builds optimized, agent-powered processes to replace them. The platform automatically notifies teams through company communication channels during implementation.
"We give you the full roadmap automatically of what needs to happen step by step for you to actually implement this agent successfully in an enterprise environment, which is often very complex," Rapoport explained. "We give you a step by step guide. 'Remove these duplicates. Connect to this data source.' And then you click on 'build' on each task, and June starts building it for you in the organization."
Real Results: CMG's Mortgage Platform Success
When CMG first considered piloting June, Akinmade was skeptical. "If your product requires FDEs, I don't want your product," he told Rapoport. "I've already done that and I'm getting annoyed by it. I don't want a black box. I don't want something only certain people can figure out. I want an easy-to-use tool."
June cleared the bar. The platform gave CMG a clear view of where to deploy agents and let them do so safely, even before the official kickoff call between the two companies. Akinmade's team had been struggling to integrate Claude Code with Salesforce — a problem that kept them from hitting their target of 100 agents. June changed that.
What This Means for Security & Compliance Analysts
For security & compliance analysts, the implications are significant. When companies rush to deploy AI agents across legacy systems without proper governance, they create security vulnerabilities. June's approach of scanning existing systems and providing step-by-step implementation roadmaps could actually help organizations maintain better oversight of their AI deployments.
The platform's ability to notify teams through existing communication channels during implementation also helps with compliance tracking. Security & compliance analysts can monitor agent deployments in real-time, ensuring that security protocols are followed throughout the process.
For more on AI security and governance, see Google's Agentic Defense Playbook and The Machine Accountability Gap: Governance for Autonomous AI Systems.
As Rapoport sees it, June complements FDEs and consultants — but her customers may be drawn to it for the opposite reason: it lets them avoid FDEs altogether. Whether that's a feature or a bug depends on your perspective. But one thing's clear: the AI deployment problem isn't going away, and companies need better tools to solve it.
