Serverless Winds of March
March 2026
Happy Spring everyone! I hope people are starting to enjoy the warmer weather in parts of the nation. San Francisco is notorious for maintaining the same basic temperature year round so it’s been more of the same for me.
KubeCon EU 2026 took place this month in Amsterdam. I wasn’t able to go but I did get some news here and there. While there is a small serverless part of the CNCF, it is largely Kubernetes based.
To me that’s fine as I don’t see them as opposing forces. I have often argued that Kubernetes CAN be serverless but people like to treat them as if they are two different things. This article seems to position them as separate options. On my regular blog I make an argument for using Kubernetes and leveraging the architecture to build a serverless platform.
Ultimately, I think for platform engineers, learning about Kubernetes and the cloud native landscape can be beneficial when building a serverless platform. So I encourage you all to spend some time reviewing some of the great talks from KubeCon EU 2026.
So onto the newsletter. I am just going to start by saying, Databricks has appeared in the news a lot this month. I personally don’t mind it. They are a great company that has helped revolutionize the serverless cloud by showing that serverless can also apply to data, not just compute. But I figured I’d give you a heads up just in case you were wondering.
Again, AI is dominating the serverless news cycle. A lot of the news seems to revolve around serverless GPUs, agent platforms and what not. However, this is also helping grow the serverless market and driving more innovation. So let’s take a look.
Serverless Industry
Serverless Architecture Market Outlook: Cloud-Native Development and Scalable Computing Growth Opportunities
Scaling Without Friction – How Serverless Architectures Are Rewriting Enterprise Growth
Startups
Product Launches
Nscale to Deploy NVIDIA Vera Rubin Platform in 2027, Bringing 100,000+ GPUs to Europe
Featherless launched a managed OpenClaw and bids to kill hidden token tax on AI agents
Google Cloud simplifies serverless security with Cloud Run’s IAP Integration
Runtimes
Nebius launches AI Cloud 3.5 with serverless AI compute
Data
Serverless Isn’t Just for Compute Anymore. It’s Coming for Your Data Stack
The Evolution of Data Engineering: How Serverless Compute is Transforming Notebooks, Lakeflow Jobs, and Spark Declarative Pipelines
AI
Tensorlake makes it simpler to deploy and scale agentic workflows with its serverless infrastructure.
Databricks rolls out Qwen3 embedding model
Databricks introduces AI Runtime: Scalable, Serverless NVIDIA GPUs on Databricks for Training and Finetuning
Learning
Efficient Serverless Container Management for Drastic Memory Reduction
App Modernization Best Practices in a Serverless World
Closing Thoughts
I think that Databricks will be a company to watch. They are still privately held but they seem to be on the bleeding edge when it comes to serverless data.
Agentic AI is continuing to push companies to look at serverless solutions. Companies like Tensorlake are leveraging the serverless architecture in order to simplify the management of agents.
We are seeing people get creative with serverless compute. Nebius is one of the largest “neoclouds” and they are launching an improved serverless AI Cloud. Dash0 is looking to improve serverless observability.
Stay tuned next month and we’ll share more serverless stories. Also keep up with my newsletter! Talk to you later!
—Photo courtesy Çağlar Çarkacı on Pexels—