In his upcoming keynote at KubeCon NA, Joseph Sandoval will discuss Kubernetes’ decade-long journey and the path forward for the cloud-native community. With Kubernetes reaching its 10-year mark and the CNCF not far behind, Sandoval aims to highlight the evolution of roles—from DevOps to SRE to platform engineering—and tools like Argo and GitOps that continue to shape the ecosystem.
A key theme in Sandoval’s keynote will be Kubernetes’ growing role in AI and ML workloads. Projects like Slurm are advancing high-performance computing in cloud-native environments while companies leverage AI to enhance platform reliability and scale efficiently. These efforts reflect how Kubernetes and CNCF projects adapt to meet end-user needs in AI-driven scenarios.
In this episode, you’ll learn about:
- The past and future of Kubernetes: Highlights from its 10-year history and what’s next for cloud-native.
- AI and ML integrations in cloud-native: Emerging projects like Slurm are bringing AI capabilities to Kubernetes.
- The role of CNCF working groups: How the AI/ML and Platform groups help drive innovation.
- End-user insights through the CNCF advisory board: How end-users provide feedback to shape CNCF project alignment.
- Resources for KubeCon attendees: Tips to get involved, from the Project Pavilion to the “Maintainer Monologues.”
This episode is sponsored by OVHcloud.
Bart Farrell (00:01.376)
It’s not every day that I get to speak with a keynote speaker for KubeCon, but today I’m joined by Joseph Sandoval. You’ll be giving a keynote at KubeCon. Can you give us any hints—without spoilers—on what your talk will cover?
Joseph Sandoval (00:13.346)
I’m speaking on Friday, which wraps up a big milestone—10 years. We started with celebrations for Kubernetes’ 10-year anniversary, including an event at Google that was amazing, full of stories and history. So on Friday, we’ll look back at these 10 years but also look forward to the next 10. A year after Kubernetes launched, CNCF was founded, so they’re also approaching a big milestone.
It’s a good moment to ask, what’s working well in the cloud-native community, and what’s missing? Over the past decade, we’ve moved from DevOps to SRE and now to platform engineering. We’ve added tools like Argo and GitOps. It’s been a continual evolution, and there’s never been a point where we could say, “Alright, everything is built—no more innovation needed.” I want us to take a look forward, especially as AI and ML continue to grow, and consider how Kubernetes will remain the platform to support these advancements.
Bart Farrell (01:56.441)
Definitely. At the last KubeCon, we spoke on a panel about the growing role of AI in the cloud-native space. You’re also speaking on a panel about cloud-native and AI integration, right? Can you talk about some of the CNCF projects that are making strides in incorporating AI?
Joseph Sandoval (02:29.55)
Absolutely. We’re seeing growth and maturity in cloud-native AI projects. At KubeCon, we’ll see projects like Slurm, which brings high-performance computing into Kubernetes clusters, supporting AI and ML workloads. For me, as part of the Adobe platform team, efficiency and scaling are top priorities, especially with diverse workloads like AI and ML.
Over the past few years, projects like KGPT have emerged. We’re also seeing an ecosystem of vendors tackling SRE and DevOps issues, bringing AI and ML into these areas. I’m excited to see more of these advancements at the event.
Bart Farrell (04:17.063)
It’s a fascinating area, especially considering the real-world applications for end users like Adobe. You’ve been involved in the AI working group for some time. For people interested in contributing, what skills are needed, and what resources would you recommend?
Joseph Sandoval (04:38.222)
This is a great time to get involved. CNCF has two active working groups: the Platform Working Group, which is very active and welcomes end-user feedback, and the AI/ML Working Group, with subgroups focused on runtimes and other areas. There are also ways to contribute to CNCF’s landscape projects.
For anyone wanting a lighter commitment, projects like the Project Pavilion tours are excellent for introductions. Getting to know maintainers this way can make involvement much easier. And for those interested in deeper insights, there’s a talk I’d recommend called the Maintainer Monologues. It’s a great resource for anyone looking to get involved.
Bart Farrell (06:46.561)
Those are fantastic suggestions—there’s something for everyone, whether they’re new or experienced. And if you’re feeling unsure, remember that the CNCF community is here to help. Lastly, I understand you’re now part of the End User Technical Advisory Board (TAB). Can you explain what that is?
Joseph Sandoval (07:14.946)
Yes, the End User TAB was formed about a year ago to represent end-user voices in the CNCF. While the TOC (Technical Oversight Committee) brings in new projects, the End User TAB provides feedback to help projects grow and adapt to real-world needs. It includes representatives from organizations at different levels of CNCF membership.
One of our first initiatives is a reference architecture to show how our platform is built, with a mix of open-source and proprietary components. It’s a way for us to share knowledge and help other end users get value from the CNCF ecosystem.
Bart Farrell (09:03.415)
It’s inspiring to see the work being done to support the community. Joseph, it’s always a pleasure speaking with you. I hope we get a chance to meet for more than 30 seconds in Salt Lake City! Thank you for all you do. We’ll see you soon.
Joseph Sandoval (09:22.766)
Thanks, Bart. Looking forward to it.
Bart Farrell (09:24.566)
Take care.