What a difference two or three days make.

When I first landed in Las Vegas for re:Invent and saw the keynote previews, it looked like AWS was putting all its chips on one square: Agentic AI. Every sign, every teaser, every whisper pointed in the same direction. And yes, after the keynotes, that impression held up. AWS is clearly betting big on agents, new model families, specialized hardware and everything needed to power the next generation of AI workloads.

But that’s only the surface story.

After spending real time on the floor, walking the expo, talking with attendees, asking vendors what they were hearing and watching the sessions closely, a very different picture emerged. DevOps, cloud native and platform engineering were not pushed to the sidelines. They were present, energetic and in many cases central to the conversations happening away from the big stage lights.

In other words, the foundations of modern tech are not quietly fading into the background. They are very much alive here. And they are thriving.

The Headlines Say “AI,” but the Work Still Says “DevOps”

There is no escaping the fact that this re:Invent leaned harder into AI than any I have attended. Frontier agents, expanded Nova 2 models, new silicon, AI orchestration systems and “AI Factory” concepts all took center stage. AWS wanted the world to know it is serious about leading the next phase of AI adoption.

What surprised me, though, was how often DevOps showed up once you looked past the main stage.

New automation capabilities appeared throughout the announcements. AWS previewed workflow and coordination enhancements for building reliable, long-running, multi-step processes. That is classic DevOps territory. The new AWS DevOps Agent landed as one of the most practical reveals at the show. A managed CI/CD runner that eliminates the need for teams to maintain their own build systems is not about hype. It is about helping organizations ship software safely and repeatedly.

And look across the expo floor. The overwhelming majority of DevOps vendors came in with force. Budgets might be tight, but the booths were busy, the conversations were serious and customers still wanted to talk pipelines, testing, automation, security, observability and tooling integration. Everyone I spoke to said the same thing: the AI wave is real, but software still has to get built. Enterprises still need to deliver reliable services. DevOps continues to be the engine that makes that possible.

Cloud Native: The Modernization Path Is Still the Preferred Route

The cloud native story is just as strong here. AWS quietly reinforced the importance of containers, serverless functions, orchestration layers and managed services across a wide range of sessions. No matter how futuristic the AI narrative becomes, modernization and digital transformation still rely on cloud-native principles.

That was the underlying rhythm of the show. Not loud, but steady and constant. You could feel it in the types of sessions that filled up. You could see it in the partner announcements. You could hear it in the way customers described their real challenges. Moving to modern architectures remains the path forward and AWS knows it.

AI may open doors to new possibilities, but those possibilities won’t materialize without the scalability, resilience and operational discipline that cloud native brings.

Platform Engineering Steps Into Its Maturity

Another encouraging sign this year is how Platform Engineering has settled into its rightful place. A few years ago, it was still treated like an emerging trend or a niche approach practiced by the avant-garde of tech organizations. That era is over.

Today Platform Engineering is accepted, well understood and viewed as an essential function for any company that wants to move with speed and safety. I had dozens of conversations with platform teams building internal developer platforms, unified self-service systems and automation layers that support both traditional workloads and the new AI-driven ones.

Organizations have realized something important. AI is not going to simplify their infrastructure. It is going to make it more complex. And that complexity needs structure. It needs guardrails. It needs consistency. All of that points straight to Platform Engineering.

This re:Invent made that truth visible. Platform Engineering was no sideshow. It was woven through the customer stories, vendor discussions and architectural sessions. It is now part of the fabric of how enterprises operate in the cloud.

Shimmy’s Take: AI Is the Future, but DevOps, Cloud Native and Platform Engineering Are the Present

Let me be clear. AI deserved the spotlight it received. The momentum is real and the innovations are coming fast. Agents, new models, new chips and new abstractions will reshape the industry.

But none of this will work in a vacuum.

AI depends on everything we have built over the past decade. It runs inside pipelines, containers, clusters and platforms. It needs observability. It needs automation. It needs governance and testing and security. It needs reliable deployment practices and scalable infrastructure. It relies on all the methods and systems that DevOps and cloud native introduced and refined.

And those systems don’t just magically appear. They are built and maintained by people who understand how to make large, complex environments operate at scale. Platform engineers, SREs, DevOps teams and cloud architects remain the backbone of modern IT.

If anything, AI makes them more important, not less.

So while the headlines shout AI, the real story happening in the halls and expo floor is this: the foundational disciplines of modern technology are still leading the way. They are the enablers. They are the stabilizers. They are the folks who make the magic possible.

It was good to see DevOps, cloud native and platform engineering so well represented at re:Invent this year. It tells me that the industry hasn’t forgotten the fundamentals. It tells me that the future we are rushing toward will still be grounded in the hard-won lessons of the past decade. And it tells me that the people doing the real work are still moving the ball forward.

AI may be hot, but the foundations are solid. And solid foundations win every time.