Another Cloud Field Day is in the books, and I still remember my first time in Silicon Valley for Cloud Field Day 2. That was the era of “cloudify everything”! The industry was moving fast, pushing workloads into environments where infrastructure was increasingly abstracted behind APIs. With just a few “clicks”, you could spin up instances running on top of some of the most advanced infrastructure ever built, operated by companies like Amazon, Google, and Microsoft.
And now? Ten years have passed, and things are evolving again. AI is now part of the conversation everywhere, bringing incredible innovation, but also a massive increase in global resource consumption. At the same time, many organizations have reached a new level of awareness: not every application belongs in the cloud.
Cloud, DevOps, and Platform Engineering
From 2016-2021, I heard a lot of claims by the big cloud providers that it is possible to migrate everything to the cloud, forgetting the effort of maintaining physical hardware. Then, in the DevOps era, everything would be possible in the cloud. So, if you reach the capacity, you can easily scale horizontally and vertically.
It was a bit like giving the keys to a candy shop to a hungry kid! In the rush to stay aligned with progress and remain competitive, companies of every size jumped into the cloud with unrestrained enthusiasm. Suddenly, everything had to be containerized, Kubernetes clusters started popping up everywhere, and governance often took a back seat.
The new motto seemed to be: “Ask for forgiveness rather than permission.”
But the real problem wasn’t the cloud itself. The real problem was assuming that moving an application automatically modernizes it. Definitely, this was the beginning of the failure of many companies that chose to move VMs and development directly to AWS, Google Cloud, and Azure without taking care of the nature of the application itself: are the “seasoned” monolithic applications ready to run in a public cloud?
If the application becomes unstable, just buy more resources. Spread the same non-optimized monolith across regions and availability zones and keep scaling vertically and horizontally. The cloud made this strategy dangerously easy: when architectural problems appear, just add more compute. Stability improves… but so does the bill.
Meanwhile, the DevOps wave pushed the accelerator even harder. Teams were encouraged to move fast, break things, and deploy everywhere. Kubernetes clusters multiplied, CI/CD pipelines exploded, and a generation of “unicorn” architectures appeared overnight. In many organizations, governance disappeared faster than the technical debt accumulated.
This is where Platform Engineering started to emerge: not as a revolution, but as a correction! Internal platforms brought back some structure to the DevOps chaos: standardized environments, reusable infrastructure patterns, and guardrails that allow developers to move fast without turning every project into its own and unmanaged infrastructure environment.
Cloud Repatriation
After years of enthusiastic migration, many companies started moving parts of their workloads back on-premises or into more controlled environments. Not because the cloud failed, but because the initial strategy was often naive.
The cloud remains an extraordinary tool, but today, the challenge is not “cloudify everything,” but finding the right balance between cloud, platforms, and sustainable infrastructure choices. Old applications could start their refactoring process on-premises, then expand their functionalities and availabilities in the cloud (Hybrid Cloud vision still matters!)
Following the recent wave of cloud repatriation, traditional infrastructure vendors are re-emerging with a renewed role, repositioning storage, compute, and networking around a simple idea the industry briefly forgot: cloud is still infrastructure… but with the right enablers it becomes a Private Cloud. Particularly interesting is the value that solutions like VMware Cloud Foundation (VCF) can provide, giving the same operational consistency that everyone has come to expect from public cloud.
Looking back at Cloud Field Day 2 after nearly a decade, it’s interesting to see how the conversation has matured. Organizations are learning to balance cloud, platforms, and infrastructure economics, understanding that architecture, governance, and operational discipline matter as much as the technology itself. Cloud didn’t disappear, of course… but it’s evolving in the direction of a more effective way to correct the excesses, and in the search for the next equilibrium between innovation, sustainability, and cost.