Platform engineering teams are broadening their focus as organizations look for consistent ways to run artificial intelligence systems. New agent-focused tools, including registries, are emerging to help teams manage AI more safely and predictably.

Early in the AI moment, tools were baked directly into an agent’s code, forcing developers to hardwire which tools an agent could use. The downside was clear: Every reuse required copy-pasting, and sharing tools across an organization was nearly impossible because everything was embedded in the code, according to Idit Levine (pictured), chief executive officer of Solo.io Inc. That’s exactly the gap Model Context Protocol aims to close for good.

Solo.io Inc’s Idit Levine talks with theCUBE about how platform engineers are meeting the maturing AI landscape.

“That’s where MCP came, and MCP basically said, ‘Listen, first, let’s standardize that because everybody should talk the same protocol,” Levine said. “If there will be a protocol, it’ll be easy.”

Levine spoke with theCUBE’s Rob Strechay at the KubeCon + CloudNativeCon NA event, during an exclusive broadcast on theCUBE, SiliconANGLE Media’s livestreaming studio. They discussed how platform engineering teams are taking on a growing role in operationalizing AI and how new agent-focused technologies are emerging to enable it. (* Disclosure below.)

Platform engineering teams in focus

MCP has created a common standard that makes it much easier for all teams to collaborate. By making tools hostable and shareable, it opened the door for anyone to use them, marking a major step forward, according to Levine. But that shift is also exposing the limits of the early AI operating model.

“At the beginning all those organizations had an AI team — they got all the money, they basically needed to build those agents and run them in production,” she explained. “But they were Python engineers; they never ran anything in production.”

As organizations move from tests toward real delivery, more responsibility for AI operations is shifting from specialized AI teams to platform engineering groups. This transition began with MCP and is now expanding into areas such as registries and agents, which will shape much of the conversation in the year ahead, according to Levine.

“I think that when people will start going to production, efficiency is going to be a big thing,” she said. “This is why skill is so important.”

Here’s the complete video interview, part of SiliconANGLE’s and theCUBE’s coverage of the KubeCon + CloudNativeCon NA event:

(* Disclosure: The Cloud Native Computing Foundation sponsored this segment of theCUBE. Neither the Cloud Native Computing Foundation nor other sponsors have editorial control over content on theCUBE or SiliconANGLE.)

Photo: SiliconANGLE