Marc Benioff-backed startup aims to solve AI deployment challenges

Big enterprises face such steep challenges in deploying reliable AI tools that entire teams of forward-deployed engineers—specialists embedded within client companies to operationalize AI systems—are emerging to bridge the gap.

“Paradoxically, AI drives up demand for professional services,” notes Efrat Rapoport, a former Salesforce executive whose newly launched firm, June, exited stealth mode on Monday. “The industry’s solution to AI integration is simply to hire more and more people.”

Rapoport and her three co-founders—Ohad Hen, Barak Goldstein, and Idan Tsitiat—propose an alternative approach to scaling AI adoption. To support this vision, the company secured $20 million in pre-seed funding led by Marc Benioff’s Time Ventures, with additional investments from tech leaders including Michael Dell, Aaron Levie, and George Kurtz. The firm declined to disclose its valuation.

The four founders previously established Bonobo AI, a pre-transformer language model startup that launched a voice-to-text service in 2017. Bonobo AI was acquired by Salesforce two years later, and the team spent several years developing the tech giant’s AI initiatives before departing after observing customers’ difficulties in integrating AI into their existing platforms.

According to Rapoport, their potential was so evident to investors that “we didn’t even need a pitch deck for this round.”

While the so-called SaaSpocalypse has software firms fearing AI replacement, no one is currently vibe-coding a CRM for a Fortune 500 company. Any AI model deployed in a corporate environment must still integrate with Salesforce, ServiceNow, DataBricks, Workday, or other dozen data-management platforms.

“Before AI can generate value, someone must address legacy systems,” Rapoport explains. “Data is fragmented across these platforms, workflows are complex, and technical debt has accumulated over years.”

Creating an agent template is straightforward, she notes. The real challenge lies in making it function amidst underlying complexity. “How does an agent operate when you have ten duplicate database fields conveying the same information, with different teams using them?”

June’s platform scans a company’s existing infrastructure to map business processes, identify bottlenecks, and construct optimized, agent-driven workflows to replace them, automatically notifying relevant teams via internal communication channels.

“We automatically generate a complete roadmap detailing the step-by-step actions required to successfully implement an agent in a complex enterprise environment,” Rapoport states. “We provide a step-by-step guide: ‘Remove these duplicates. Connect to this data source.’ Then, you click ‘build’ on each task, and June constructs it within your organization.”

Paul Akinmade, chief strategy officer at CMG, a major U.S. mortgage lender, rapidly shifted his company’s software engineering to Claude Code but encountered integration hurdles with Salesforce. This posed a problem, as he had promised at Salesforce’s annual conference the previous year that he would return with 100 active agents, a target now appearing unattainable.

Akinmade reports his team spent weeks hitting roadblocks—consulting architects, engaging forward-deployed engineers, and seeking advice from everyone available—without progress. June resolved this, he says, providing his team with clear visibility into agent deployment locations and enabling safe implementation, even before the official kickoff call between the two companies.

Rapoport views June as a complement to FDEs and consultants, yet her customers may prefer it for the opposite reason: it allows them to bypass FDEs entirely.

When Akinmade first considered piloting the tool at CMG, he told her: “If your product requires FDEs, I don’t want it. I’ve already experienced that and found it frustrating. I don’t want a black box. I don’t want something only select individuals can navigate. I want an easy-to-use tool.”

Clearly, June met that standard.