Red Hat AI-ifies OpenShift; JFrog finds "shadow AI"

Welcome to Runtime! Today on Product Saturday: Red Hat rolls out the latest version of OpenShift at KubeCon, JFrog's new tools helps companies understand how they're using AI, and the quote of the week.

(Please forward this email to a friend or colleague! And if it was forwarded to you, sign up here to get Runtime each week.)

Before we get into the roundup, allow me to thank everyone who became a Runtime Supporter this week. For just $10 a month, you're helping us bring independent enterprise tech journalism to the people who need reliable, quality information to make the decisions that will shape their companies. We'll roll out more information about our Supporter program in the coming weeks, and if you value the work we've done over the past three years, please consider becoming a supporter.

Ship it

Shifting gears: The annual KubeCon North America show took place in Atlanta this week, and IBM's Red Hat used the occasion to update OpenShift, its container-management platform built around Kubernetes. Version 4.20 (nice) recognizes that AI workloads require different types of care and feeding compared to traditional workloads, and is now available for Oracle's cloud infrastructure.

A new feature called LeaderWorkerSet allows customers to "efficiently distribute and scale AI models that exceed a single pod's capacity, using a leader pod to seamlessly orchestrate communication across worker pods," Red Hat said in a blog post. OpenShift 4.20 is now more adept at load-balancing AI workloads, and allows data scientists to make changes to AI models without having to rebuild the associated container with every update.

Keep it fresh: Under CEO Dennis Woodside, Freshworks is trying to take on bigger clients after years of targeting small and medium-size businesses with its business-management software. This week at its Refresh event the company released new updates to its customer-service software as well as its IT management platform.

Freshdesk (for customer service) now has a new central hub for managing ongoing customer issues, and new agents for customers in industries like fintech and travel. On the IT management side of the house, Freshservice now allows administrators to monitor the health of end-user devices running on their network in hopes of detecting issues before they become real problems, and now comes with agents that can find data stored in a Google Drive account.

Hide and seek: The lack of visibility a lot of companies have into the tools, software, and cloud services that employees are using to get their work done is a recurring problem in enterprise tech, and generative AI tech opens up a whole new can of worms. This week JFrog released a tool that promises to help customers find "shadow AI" within their environments, a nod to the longtime "shadow IT" problem.

Shadow AI Detection "helps automatically detect and create an inventory of all internal AI models and external API gateways used across the organization to access data from either approved or ad-hoc third-party sources," the company said in a press release. It also helps those customers establish an audit trail for compliance purposes.

Lake of the clouds: One thing that has separated companies struggling to deploy AI agents and companies that have figured it out is that in most cases, the latter group walked into the room with a more comprehensive and modernized data management and storage strategy. Dremio is one of several companies that has embraced the data lakehouse architecture as a vehicle for sprucing up corporate data, and this week it introduced a managed cloud service that uses agents to manage that data.

"... Dremio Cloud transforms legacy-based data infrastructures into a self-managing platform that continuously learns, adapts, and optimizes without human intervention," the company said in a press release. The new service also promises to analyze trends across data usage and label data as it is created.

No exit: Just about everybody who has ever created an account with a cloud provider loves to complain about egress fees; it's super easy to get your data onto a cloud provider's servers, but getting it out is far more annoying. CoreWeave is very eager to get traditional cloud customers onto its AI infrastructure services, and this week it introduced a migration service that makes it cheap and easy to do just that.

The Zero Egress Migration program will allow "customers to move large-scale datasets from third-party cloud providers directly to CoreWeave’s cloud without technical friction," the company said in a press release. And CoreWeave said it would cover any egress fees charged by the current storage provider and customers "will not incur any exit penalties from CoreWeave afterwards," it said.

Stat of the week

While the complexity of agentic AI is probably the largest obstacle holding back companies trying to get on board, security concerns are also top of mind. According to new research from GitLab, 76% of developers, security professionals, and operations engineers agreed with the statement that "Agentic AI will create unprecedented security challenges for our organization to navigate.ˮ

Quote of the week

"It’s a cautionary tale. We should definitely take everything they say from now on with a grain of salt, just because of how this all transpired." — DA Davidson's Gil Luria, discussing the nearly 35% drop in Oracle's stock since it announced that preposterous infrastructure deal with OpenAI in September that seems very unlikely to play out as advertised.

The Runtime roundup

SAP will likely avoid fines as the result of an antitrust investigation in the E.U. after agreeing to change some licensing practices and allow customers to get support from a wider variety of partners, according to Bloomberg.

Cognizant strengthened its position with Microsoft Azure customers after snapping up 3Cloud, which was just named Microsoft's partner of the year, according to CRN.

Thanks for reading — see you Tuesday!