I'm Loving Serverless
February 2026
February was a very interesting month for me. I got to see a massive snow storm. My girlfriend and I went skiing and it happened to be when that huge storm hit giving us around 110 inches (~282cm) of snow! I am going to tell you right now, this South Texas boy is not build for living in those conditions but we made it.
Also our hearts go out to those impacted by the lives lost during the snowstorm and the avalanche. You will be missed.
Not the best segue but maybe we can talk about some serverless stuff now. I already wrote a newsletter article on Heroku entering “maintenance mode” which, to me, was the biggest story of February. I still consider them one of the originals. While PaaS isn’t serverless per-se, it laid the foundation on which serverless was built.
But there were also a lot of acquisitions and new offerings so let’s jump in.
Serverless Industry
Serverless Compute Market set to TRIPLE!
Sunsetting the Print Server, why you should go serverless.
Startups
Mistral acquires Kyoeb, a serverless platform for models
Cloudflare acquires Astro, which may simplify webdev on Cloudflare
S2.dev raises $3.85B for serverless data infrastructure
Product Launches
Google introduces NVIDIA RTX PRO 6000 Blackwell GPUs to Cloud Run
Databricks introduces Lakebase, a serverless Postgres database for AI
AWS launches 3-year serverless reservations
Databricks launches Zerobus Ingest for serverless data streams
Azure now offers foundry models with serverless APIs
Runtimes
Cloudflare introduces vinext, a fork of Next.js that runs anywhere, in one week, with AI…
Heroku entering “maintenance mode”
Data
Simplyblock releases a branchable serverless Postgres database
AI
Komprise makes Agentic AI with Serverless Compute easier
Six companies launch Project DI, a global AI serverless platform.
Nebius publishes an outlook on Serverless AI in 2026
Learning
Cloudflare shares a case study with TightKnit
Hidden Serverless security risks you should look out for
Closing Thoughts
You may have noticed that this month was short on stories related specifically to runtimes. That’s actually a great thing as it shows that serverless is evolving not just beyond functions but also beyond compute runtimes. People are thinking about running agents on serverless, using serverless for data streaming and processing, and even storing data with serverless databases.
Serverless is more than just functions and compute. It can be used for a lot of data and I believe it will be a bedrock for AI applications moving forward. We are already seeing a lot of investments for serverless platforms and agents THRIVE on serverless.
Now I know OpenClaw was a huge story recently and you may be wondering how/if that plays into serverless platforming. I am avoiding writing about it because 1) everyone else is so I don’t know what I can add and 2) I am waiting for the dust/hype to settle to fully evaluate it. I haven’t tried running OpenClaw on Cloud Run yet but I may give Nano Claw a try and report back.
Join me in March for the next summary newsletter and stay tuned for my bi-weekly updates. See you soon!
—Photo courtesy Karley Kosmos on Pexels—