In an uncanny echo of pioneering Tintri's Alexa speechbot work, private cloud supplier and VMware alternative VergeIO is enabling AI agents to work directly with a customer’s VergeOS environment, building networks, deploying workloads, and diagnosing faults in plain language.
The company has announced two related items. Verge CLI is a complete command-line interface for VergeOS and an MCP server built on the open Model Context Protocol with a set of agent skills. Used together, they let agentic AI platforms, including Anthropic’s Claude Code and OpenAI’s Codex, manage and monitor a customer’s VergeOS environment.
VergeIO’s Larry Ludlow, the Chief Architect of Verge CLI, said: “The agent reasons against VergeOS documentation through the MCP server, so its diagnoses come from how the platform actually works, not a model’s guess. Because one API spans compute, storage, and networking, it traces a fault across the whole stack that tooling stitched across separate products would miss.”
The Verge CLI maps to the full VergeOS API, so one command set covers compute, storage, networking, and data protection. That command set is what an AI platform uses to act. A set of VergeIO agent skills provides the means to drive the command set. Claude Code or Codex reads the environment, proposes the work, and runs it within the limits the administrator sets.
Infrastructure that used to mean per-core VMware licensing now takes direction in plain language.
Jason Yaeger, SVP of Product and Engineering at VergeIO, said: “Customers leaving VMware want lower cost and infrastructure ready for AI. Verge CLI gives them both. Claude Code, Codex, or a local model they run themselves can all operate the platform directly, and the administrator stays in control of what it’s allowed to do.”
Because the agent interface is via MCP, any MCP-compliant model or agent can be used. Users with privacy or security requirements can run a local open-weight model, such as Llama, Qwen, or DeepSeek through a runtime like Ollama, and keep every operation and all environment data on their own infrastructure.
Comment
Nine years ago storage supplier Tintri, which was focussed on storage specifically for VMware, blazed a trail for an AI interface to a storage array’s management functions. It developed its Alexa speechbot, whereby the Amazon Echo’s Alexa speech recognition SW could be used to trigger array system management operations.
In a YouTube video, Tomer Hagay, Tintri’s director of technical marketing, tells Alexa to provide the status of a system, provision three VMs, and then report on the operational status of an array. And that is what the speechbot did.
As we wrote at the time, the video starts with him saying: “Alexa; ask Tintri what it can do?”
Alexa replies (in speech) “Provision VMs. Take snapshots. Apply QoS. Add production data to VMs. Provide real-time stats and many more.”
Hagay then tells Alexa to provision three VMs and receives spoken, vCenter display, and email confirmation that it’s been done. And Alexa managed to have the Tintri array SW carry out these tasks. An interview with Tintri CEO and then CTO Kieran Harty, tells you more about it.
This Tintri capability was derided at the time but here we are, with natural language text input to systems like VergeIO. Generative AI chatbot interfaces to management operations are becoming normal. It will surely not be long before speech interfaces get added as well.
As for Tintri, it was subsequently bought by DDN after it nearly went bankrupt, and, presumably, somewhere in the DDN-owned Tintri assets vault, the Alexa speechbot interface languishes, a now silent and forgotten pioneer of what is rapidly becoming an everyday occurrence.