June's $20M says AI deployment is a legacy systems problem

A startup called June just raised $20 million in pre-seed funding — an absurd number for a company with no product in general availability — to solve a problem nobody wants to admit they have. Not "how do we build AI agents." AI agent deployment fails at a duller step: finding out where your data actually lives and which process is worth automating in the first place. June's pitch is that this discovery work is itself an AI problem.

What actually happened

Per TechCrunch, June's round was led by Marc Benioff's Time Ventures, with Michael Dell, Aaron Levie, and George Kurtz also participating. The founding team — Efrat Rapoport, Ohad Hen, Barak Goldstein, and Idan Tsitiat — previously built Bonobo AI, which Salesforce acquired in 2019. That history explains both the check size and the target: they know exactly how ugly the inside of a Salesforce org gets.

The product scans a company's existing systems to identify its real business processes and where they bottleneck, then generates an implementation roadmap and walks the company through integrating agents with legacy platforms — Salesforce, ServiceNow, Workday. Rapoport's framing, per TechCrunch: "Before AI can create value, someone has to deal with legacy systems."

The named customer is CMG, a US mortgage lender that had been stuck on Salesforce integration. CMG chief strategy officer Paul Akinmade told TechCrunch the tool gave the team a clear view of where to deploy agents, and let them deploy before the official project kickoff.

Why it matters for your business

Strip out the enterprise scale and this is the exact failure mode we watch small operators hit. A business decides to "add AI," picks a tool, and then spends four months discovering that the customer record in the CRM disagrees with the one in the billing system, that the process everyone described in the kickoff meeting isn't the process anyone actually follows, and that the highest-value automation was three steps upstream of where they started building.

The reason June can raise $20M pre-seed on this is that the discovery step is genuinely the hard part, and it is genuinely underserved. But notice what the product produces: a map and a roadmap. That's a consulting deliverable wearing software. You don't need a funded platform to get one — you need someone to trace your actual workflows, find where data goes stale or gets re-keyed by hand, and rank the automations by payback. If a vendor tells you the model is the hard part, they haven't seen your systems.

The other lesson is in the customer story. CMG deployed before the official kickoff. That's the right shape: prove one narrow workflow works, in production, before anyone budgets a program around it.

Sources: TechCrunch.