Nutanix doubles down on agents; Zencoder branches out

Welcome to Runtime! Today on Product Saturday: Nutanix hopes to entice more VMware customers to make the jump with support for AI agents, Zencoder wants to help developers automate the non-coding parts of their day, and the quote of the week.

Please forward this email to a friend or colleague! If it was forwarded to you, sign up here to get Runtime each week, and if you value independent enterprise tech journalism, click the button below and become a Runtime supporter today.

Ship it

Agents, converge: More than two years after Broadcom acquired VMware, recent research found that 86% of North American VMware users would like to go in a different direction. Nutanix has been courting those users ever since that transition closed, and this week introduced new AI capabilities for its flagship cloud service that could help nudge them along.

Nutanix Agentic AI, which was announced last month at Nvidia GTC, is now available through an early access program for Nutanix Cloud Platform customers. It's "a full-stack AI solution to go and offer full access to large language models that are curated and certified, running on certified GPUs, including an AI gateway," Nutanix's Lee Caswell told CRN.

Circular file: Countless businesses were built on the back of AWS's S3 object-storage service, which kicked off cloud computing as we know it today. But agentic AI workloads need to be able to work with files to execute tasks, and this week AWS added native support for file storage to the 20-year-old service.

Agents can now access data stored in S3 buckets as if that data was stored locally on a computer's file system, which eliminates the workarounds needed to make agents work with object data. "By making data in S3 immediately available, as if it's part of the local file system, we found that we had a really big acceleration with the ability of things like Kiro and Claude Code to be able to work with that data," AWS's Andy Warfield told VentureBeat.

Make it work: Coding agents like Claude Code and Kiro (which, unless you work at Amazon, you might not have tried) have quickly become part of software development workflows over the last six months, and they're reminding developers that coding was always just one part of their job. This week Zencoder introduced Zenflow Work, a new addition to its own coding orchestration service that helps developers prepare for meetings and work up documentation.

Developers are already using coding agents for those kinds of projects, but "the problem is that the interfaces were designed for engineers who are comfortable with Git, PRs, work trees, and command-line configuration," which makes them hard to use when working with non-engineers,  the company said in a blog post. "That’s one piece that’s missing from a lot of AI tools right now: the ability to not just collaborate with your agent, but collaborate across the team with the agents,” CEO Andrew Filev told The New Stack.

Middle management: Given that every enterprise software company with a pulse is trying to build the middleware that will help companies build and deploy AI agents, it's not too surprising that Anthropic would eventually get around to jumping into that space as well. This week it introduced Claude Managed Agents, a new service that resembles an application-development platform geared around agents instead.

"Managed Agents pairs an agent harness tuned for performance with production infrastructure to go from prototype to launch in days rather than months," Anthropic said in a blog post. Once customers lay out the details of what they want their agents to accomplish, Anthropic's software will deploy those agents on its infrastructure and manage any issues that pop up.

Context contest: ServiceNow is definitely one of the companies trying to build an agentic AI layer over its core products, and last year at ServiceNow's Knowledge 2025 event CEO Bill McDermott suggested that the best way to do that was to connect different lines of business across a company with the right data and tools. This week ServiceNow introduced Context Engine, which promises to do just that.

Context Engine "gives ServiceNow AI and workflows the context to sense what’s happening across the enterprise, decide the right course of action, act with precision, and govern every outcome accountably," the company said in a press release. There is, of course, another angle: "Most enterprise scale SaaS providers are feverishly incorporating AI functionality horizontally across their solutions, all seeking to lock their customers in leveraging their own flavors of data fabric and context layers," Info-Tech Research Group's Scott Bickley told CIO.

Stat of the week

Last November's crop of new AI models jump-started the AI coding agent race and allowed enterprises to start taking agentic AI more seriously, but a year's worth of empty promises about AI agents has a lot of companies still in wait-and-see mode. According to new research from Writer, while 75% of tech leaders think agents will play a significant role in how they manage their company within the next five years, only 23% have seen "significant ROI" from agents and 48% think their AI efforts to date have been "a massive disappointment."

Quote of the week

"We basically need to start, right now, preparing for a world where there is zero lag between discovery and exploitation.” — Anthropic's Logan Graham, on the consequences of rapid advances in AI model technology when it comes to finding vulnerabilities in software and — more importantly — immediately taking advantage.

The Runtime roundup

CoreWeave's stock closed up nearly 11% Friday after it announced a multiyear agreement to provide infrastructure services for Anthropic.

Meanwhile, Anthropic is thinking about designing its own chips, according to Reuters, which would make the anything-but-Nvidia movement even more interesting.

As Project Stargate continues to fall short of the lofty expectations it once had, three OpenAI executives working on the project are planning to leave the company, according to The Information.

Thanks for reading — see you Tuesday!