FinOps, AI-Driven Operations, and Next-Gen Observability Take Center Stage at KubeCon 2025
Alastair Cooke’s KubeCon 2025 takeaways highlight a shift from simply building Kubernetes to efficiently operating it at enterprise scale, with FinOps and cost optimization moving to the forefront through smarter autoscaling, tooling, and hybrid infrastructure reuse. AI’s role has become more grounded, with practical SRE-assist features, integrated AI tooling across the app lifecycle, and security gaps increasingly addressed by network security layers and proxies around fast-evolving AI platforms. Meanwhile, observability has resurged in importance thanks to OpenTelemetry’s rise as a unifying standard, enabling richer, AI-powered insights to keep pace with the growing complexity of Kubernetes environments.
Transcript
I'm Alistair Cook, an event lead here at Tech Field Day, and these are my takeaways from KubeCon in 2025. Tech Field Day standalone feels very different from Tech Field Day at a conference. And so there is, uh, an impression that is much wider than just what we saw on the stage at Tech Field Day.
I saw a lot of observability being called, again, I saw a lot of threads of AI right through all of the products that are being delivered. And what we're seeing is really very much a maturity that the people who are coming to CubeCon now are about running infrastructure in production separate from actually building the platform itself. And I think that's very much a maturity story around Kubernetes as just another vital infrastructure platform.
And we need all of the tooling around Kubernetes that we've always needed around our on-premises and cloud infrastructure. My first big takeaway is that ops and cost optimization is getting a lot more focus as Kubernetes is now a production platform. We're running workloads on Kubernetes long term, and we're thinking about how to make sure we get the most value out of those applications without having excessive spend on the infrastructure underneath them.
So we're looking at large scale deployments, both in public cloud, but also increasingly on premises alongside existing applications. And where you've got that mix of legacy applications and newer applications, maybe VM based client server applications alongside Kubernetes based cloud native applications. That complexity of mixes becoming a, a really significant part of that cost discussion.
Sometimes the right places to run on premises, sometimes the right places to run on the cloud and always making sure that you're getting value for the dollars that you spend. This has become much more of a driving thing to me in Kubernetes recently. The second thing that I've seen is that AI is everywhere, but it's not that everything has been AI washed so much as it was last year.
This year, AI has become the feature rather than the focus. So new capabilities are being built into infrastructure and particularly observability platforms to use both predictive AI that statistical, how did things behave in the in the past, but also generative AI come up with some new insights and new information for me and provide an interface that's a little more friendly. Both aspects of this generative and predictive AI we're seeing both in the applications that are being delivered, but also there's a focus on AI as an application that's being delivered.
The security and the observability requirements that wrap all around these new AI applications. And we're seeing a lot more maturity in that. Speaking of observability, observability has become cool again.
Observability had a, a big rush a few years ago, went a little bit quiet because observability is tough and is starting to have a bit of a resurgence now. The core idea of observability is that we extract from our running application the information we need to know its state over time. We don't have to look inside the running application to see its state.
We've pulled it out to somewhere else and we ask questions about that somewhere else when we need to deal with performance problems or reliability problems when we're looking to improve our application. One of the challenges of observability has always been there's a huge amount of data we are pulling outta these applications and analyzing that huge amount of data is difficult. Well, this is where AI is tending to come in and help us to analyze large volumes of data because that's what computers should be really good at.
And so there was definitely been a resurgence of the, uh, observability practice and the value that observability brings to particularly fast moving, fast changing cloud native applications even when there are alongside legacy applications for the slow moving. That observability value has been a very vital and it was definitely a resurgence at Cube Comm. I really enjoyed CubeCon, I enjoyed the companies we had presenting for Tech Field at CubeCon.
I highlight for me was having traffic labs. I was briefed by them earlier in the year thought the having a story that goes beyond just the free traffic labs that many of us have used, uh, was great to have a bunch of different delegates coming and joining us as well. Uh, particularly having some of the future analysts making some time in their schedules.
And I know some of our delegates had very busy schedules at CubeCon because it is such a fun, exciting conference to be at. Thanks for watching this episode of the Tech Field Day takeaway series on the Tech Field Day YouTube plus channel. If you enjoyed it, be sure to like, subscribe and share your thoughts, uh, on the announcements from CubeCon in the comments follow Tech Field Day on X, Twitter, blue sky Mastodon, and all of your other favorite social media platforms.
And check out all of the presentation videos on the field day website and YouTube channels. We'll see you at the next field day event and check out tech field day com for details of all of our upcoming events.
