Geniez AI announced a $6 million seed round on September 4, 2025, co-led by StageOne Ventures and Canapi Ventures. Thirteen months later, Rocket Software acquired the six-person startup for a reported up to $100 million, according to CTech.
The acquisition marks one of the faster enterprise software exits in recent memory, though Rocket Software did not confirm the valuation reported by CTech when it announced the deal on October 13, 2026. What the company did acquire was an AI agent framework designed to bridge large language models with mainframe systems, a niche but lucrative corner of enterprise technology that has proven surprisingly durable.
Geniez was founded in 2025 by Gil Peleg and Dan Shprung, both veterans of the mainframe world. The pair had previously built and sold Model9, a mainframe data-management startup, to BMC Software in 2023. That track record appears to have mattered. When announcing the seed round, Yuval Cohen of StageOne Ventures pointed to "a proven track record of innovation in the mainframe space." Neil Underwood of Canapi Ventures added that AI had become essential infrastructure for financial institutions, many of which still rely heavily on mainframe systems.
A short runway to acquisition
At the time of its seed funding, Geniez outlined plans to scale engineering and expand enterprise partnerships, according to Business Wire. The company installed its framework at financial institutions and large enterprises operating mainframes, though it never disclosed customer names publicly. The speed of the exit suggests Rocket saw strategic value in absorbing the technology before Geniez reached significant scale.
CTech reported the deal valued the startup at up to $100 million for a team of six employees. That figure, if accurate, reflects either exceptional technology, competitive pressure, or perhaps both. Rocket did not confirm it.
Peleg, who led Model9 as CEO before its acquisition, has spent 25 years in mainframe infrastructure, according to BMC's 2023 announcement of that earlier deal. Shprung served as Model9's chief revenue officer and later became sales director for BMC's AMI Cloud division before reuniting with Peleg to start Geniez. In a September 2025 interview with Blocks & Files, Peleg drew a direct parallel between the two ventures: "We're going to bring GenAI to the mainframe just like in Model9 we brought object storage to the mainframe."
How the technology worked
Geniez built a framework that runs natively on mainframes and connects to large language models operating on-premises or in public clouds, including OpenAI's ChatGPT, Anthropic's Claude, and Meta's Llama. The system relied on the Model Context Protocol and agent-to-agent standards, and provided a Python SDK for developers, according to the company's funding announcement.
The startup integrated its platform with GitHub Copilot, Amazon Q Developer, and Claude Code for mainframe development tasks, according to archived snapshots of the company's website from October 2026. It also connected to Microsoft Teams, Jira, ServiceNow, and GitHub for operations workflows.
Geniez claimed its framework could monitor live mainframe systems, diagnose incidents, remediate security issues, and execute workflows through natural language commands. Installation, the company said, took hours rather than weeks. Whether that held true at scale remains unclear given the compressed timeline between launch and acquisition.
Mainframe AI heats up
Rocket Software competes with IBM, BMC, and Broadcom in the mainframe software and modernization market. IBM introduced multi-agent development capabilities for mainframes in July 2026. Rocket itself had expanded its EVA agentic AI platform in September 2026, just three weeks before acquiring Geniez, adding governance controls for agent access to mainframe systems. The acquisition filled an obvious gap.
Mainframes remain more resilient than many observers expected. A 2026 BMC survey found that 83 percent of respondents reported transaction volumes either increased or held steady on their mainframe systems. Financial institutions in particular have shown little appetite for replacing infrastructure that still handles the bulk of global transaction processing.
Milan Shetti, CEO of Rocket Software, framed the acquisition broadly in the company's announcement: "This is about continuing to transform what is possible on the mainframe." That statement, like many in corporate acquisition announcements, revealed little about integration plans or product roadmaps. What it did signal was urgency. Rocket moved quickly, and it paid a premium to do so.
