Introduction: The Paradox of Enterprise AI Deployment
According to a report by TechCrunch, massive corporations are struggling to run artificial intelligence (AI) tools reliably in their live production environments. This persistent challenge has become so pronounced that an entire niche industry of forward-deployed engineers (FDEs)—highly specialized professionals who physically drop into enterprises to deploy and stabilize AI systems—is rapidly expanding to meet the demand.
"AI, paradoxically, increases the demand for external professional services," explains Efrat Rapoport, a former senior executive at tech giant Salesforce. Rapoport's new startup, June (also known as June AI), officially emerged from stealth mode on Monday morning, August 3, 2026. "The industry's current answer to AI implementation is simply, 'let's hire more and more and more people.'"
The Funding and June’s Founding Team
Rapoport and her co-founders—Ohad Hen, Barak Goldstein, and Idan Tsitiat (Idan Tsitiat)—have a completely different vision for how to bring artificial intelligence into widespread operational use. To turn this vision into reality, the startup has raised $20 million in a pre-seed funding round. The round was led by Marc Benioff’s Time Ventures, with additional backing from prominent technology figures including Michael Dell, Aaron Levie, and George Kurtz. The company has declined to disclose the valuation at which this funding round was secured.
The four co-founders are no strangers to the AI space. They previously built Bonobo AI, a pioneering conversational intelligence company that developed pre-transformer language models and launched an early voice-to-text service in 2017. Just two years later, in 2019, Bonobo AI was acquired by the enterprise giant Salesforce. The founding team subsequently spent several years leading various AI initiatives inside Salesforce. It was during this tenure that they watched enterprise clients continuously struggle to integrate and deploy new AI technologies into their existing legacy platforms. This glaring operational bottleneck inspired them to venture out on their own once again. According to Rapoport, the sheer potential of the founding team and their core thesis was so clear to investors that they did not even need to compile a traditional pitch deck to close the pre-seed round.
The Real Problem: Legacy Systems and Technical Debt
While the industry phenomenon known as the "SaaSpocalypse" has software firms deeply worried that AI will completely replace them, the reality on the ground is far more complex. Today, no one is "vibe-coding" a core customer relationship management (CRM) platform for a Fortune 500 corporation. Any AI model introduced into a modern, complex corporate environment must still interact flawlessly with established platforms like Salesforce, ServiceNow, Databricks, Workday, or any of the dozens of other systems managing the enterprise's data.
"Before AI can create any value in an organization, someone has to deal with the legacy systems," Rapoport points out. "You have fragmented data spread across all of these different platforms. You have incredibly complex workflows. And you have years of accumulated technical debt."
According to Rapoport, building a simple agent template is actually the easiest part of the process. The real obstacle lies in making that agent operate within the chaotic data environments lying beneath the surface of the organization's legacy systems. "How does an AI agent know how to operate when you have ten duplicate fields in your database that say exactly the same thing, and different business teams are using them?" she asks.
June’s Solution: Automated Mapping and Deployment Roadmaps
To solve this, June has developed a technology platform that automatically scans an enterprise's live systems to build a deep understanding of its actual business processes. The platform identifies workflow bottlenecks and automatically designs optimized, agent-powered processes to replace older, inefficient workflows. Once deployed, the system can automatically send alerts and real-time updates to relevant teams via the organization's existing internal communication channels.
"We automatically provide you with a full roadmap detailing step-by-step exactly what needs to happen so you can actually implement this agent successfully in an enterprise environment, which is often very complex," says Rapoport. "We give you a step-by-step guide that says, for example: 'Remove these duplicates. Connect to this specific data source.' And then you just click the 'build' button on each task, and June's platform starts building the process for you within the organization."
Case Study: Deploying Agents at CMG
The power of June's approach was recently demonstrated at CMG, a major mortgage lender in the United States. Paul Akinmade, Chief Strategy Officer at CMG, had successfully transitioned his team's software engineering operations over to Claude Code, but hit severe technical roadblocks when attempting to integrate it with Salesforce. This integration was critical, especially after Akinmade had publicly promised at a major conference to return with 100 active AI agents running in production.
Faced with weeks of stagnation and fruitless meetings with enterprise architects and forward-deployed engineers, CMG turned to June. The platform quickly resolved the impasse, providing Akinmade's team with a clear, visual map to safely deploy their AI agents. Crucially, June allowed CMG to deploy the agents successfully even before the official kickoff call between the two companies.
Replacing Forward-Deployed Engineers with Easy-to-Use Tools
While Rapoport views June as a platform that can complement the work of forward-deployed engineers and consultants, her clients are increasingly drawn to it for the opposite reason: it allows them to bypass FDEs entirely. When initially evaluating the tool for CMG, Akinmade was direct with Rapoport: he did not want a black box that required specialized, highly expensive talent to configure. Instead, he demanded an easy-to-use tool that his team could operate independently. June successfully cleared that bar, proving that enterprise AI deployment can be automated rather than manually staffed.