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

• Serverless Architecture: The Future of Scalable Websites

• Scaling Without Friction – How Serverless Architectures Are Rewriting Enterprise Growth

Startups

• Dash0 Acquires Lumigo to Expand Agentic Observability Across AWS and Serverless

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—

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

• Overview of Serverless Postgrew marketplace

• 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

• Databricks finalizes a $5B equity raise.

• Simplyblock releases a branchable serverless Postgres database

• Expert canvases innovative tech for data management

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

• Cracking the code of serverless design patterns

• Interview With Voximplant Founder & CEO Alexey Aylarov

• 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—

This tells me that Salesforce is effectively shutting down Heroku. Now I don’t know why this is the case but my suspicions are that Salesforce doesn’t want to be in the hosting business anymore.

We something similar back when Facebook bought Parse in 2013 only to shut it down in 2017. It makes sense as managing a hosting platform does take a lot of additional resources and doesn’t make sense if it isn’t your core business. That being said, Heroku is an OG.

What is/was Heroku

Heroku is one of the original Platform-as-a-Service (PaaS) providers. They were founded back in 2007 (Pre-dating Google App Engine by a year). To shed some context, in the early days of the cloud, everyone wanted virtual machines (VMs). This made sense as that was the hot tech in the 2000s and early 2010s.

That being said, developers didn’t care for it. They just wanted to host code. Heroku created an abstraction layer between VMs and the developer to simplify the deployment of applications to the cloud. I actually talk a little bit about this in my other blog.

Now creating a PaaS solution wasn’t their only contribution to technology. They also created the concept of the 12-factor app. Further, they open sourced the concept in 2024 during KubeCon North America. The also walked so that serverless can run in many ways.

I don’t think we would have serverless today if we didn’t have PaaS. It was a necessary stepping stone in redefining the cloud and making it more than just hosting VMs on someone else’s server.

Fun fact, in the early 2010s I was working on a startup (that got nowhere) that was modeled after Heroku (and dotCloud which eventually became Docker). I hated using VMs and bare metal servers (my first website was hosted on a server in my office area) and figured most developers did too.

Why the Change?

I have never been in a Salesforce boardroom nor do I know why they make decisions. But I can speculate. My theory is that they simply didn’t want to put money towards Heroku.

Unless you have been living under a rock, Agentic AI has been dominating the tech space for about a year and a half or so. There has actually been a lot of talk about Agents replacing SaaS (Software-as-a-Service). By the way, Salesforce has historically been a SaaS model.

Salesforce saw the writing on the wall so they went ahead and created their Agentforce platform. To put it simply, they created a way to host and scale agents. Salesforce has pivoted their business to lean heavily into this agentic AI space.

In this AI race, they likely didn’t really see a place for Heroku in their portfolio. Sure, they could repurpose it to be an agentic platform but then it would just compete with Agentforce. In my opinion, this is why they made the decision. They want to divert resources towards Agentforce and so something had to go.

Is This the End?

While I am sad to see this happen to Heroku, this probably isn’t the end of them. They may bounce back or be absorbed into Agentforce in some way.

Outside of Heroku, we have seen an explosion of Agentic AI platforms that leverage serverless architecture. I don’t think this is the end of PaaS or serverless but just marks an evolution or “passing of the torch”.

So let’s all pour one out for one of the OGs…

—Photo courtesy Aarón Lares on Pexels—

Now again, I understand why this is the case. Prior to the launch of AWS Lambda, the term “serverless” didn’t really exist in our vocabulary. Lambda was the first commercially viable Functions-as-a-Service (FaaS) platform and AWS labeled it as “serverless”.

To give AWS credit, Lambda and by extension FaaS was the first platform to popularize the term serverless. But we really need to shake the idea that serverless is limited to just functions.

So a quick note, I am a Googler and I am a huge fan of Cloud Run. I have been supporting that product as a Googler since its beta launch in 2019. So naturally, I will always promote it. However, I am only going to focus on the facts as it relates to the post on The New Stack and not focus on other features. I will also highlight other companies doing similar things with products.

With that out of the way, let’s address the two key arguments here.

VPCs and On-Premise Data

This quote from the article sums up their stance pretty decently on this topic. “Enterprise AI programs must now assume that agent workloads touching internal data will require VPC-native or on-premises deployment.”

This makes perfect sense as a lot of internal data is behind a VPC or on-premise as it sensitive in nature. Your agent, wherever it is running, needs to access the data. But let’s take a step back. Are they arguing that serverless platforms can’t utilize VPCs? I mean, Cloud Run can. Heck, even Lambda can. You can configure a VPN to connect to on-premise data.

Serverless Data Persistence

Serverless, by nature, is stateless. The service will “wake up”, execute its task, and spin down. When it “wakes up” again, it has no real knowledge of what it did last time. This is a great architecture for workloads such as API endpoints or batch processing data.

Agents, on the other hand, require some statefulness. It needs to know what task it previously completed as it continues to performs autonomous and asynchronous tasks. Traditional serverless platforms make this a challenge.

Cloud Run does support ways to address this though. For example, you can use Memory Bank for short-term memory storage. Long term memory storage can be put into a simple key-value based NoSQL database. Lambda can leverage DynamoDB for data persistence, for example.

Ultimately, I think this article looks at serverless in a vacuum. It sees serverless as just a runtime. If you approach your cloud architecture in that fashion then sure, you will see some shortcomings and a VM may make more sense. However, with every major hyperscaler and even the neoclouds, you can connect serverless runtimes with additional services to easily overcome these issues.

Serverless architecture isn’t failing agents. It’s helping it thrive.

—Photo courtesy Brett Jordan on Pexels—

I personally continued with my tradition of skiing in Tahoe for NYE. It was a pretty fun day overall and the snow was decent enough.

Now that we are in 2026, I will do some serverless predictions.

1. More Serverless AI, in particular, around agents. I think we’ll see more and more startups and established companies create serverless-native Agent platforms. I think we will see more agent orchestration layers that leverage serverless technology.

2. Advancements in serverless beyond FaaS. With the growing popularity of WASM and the graduation of Knative, I think we’ll see more startups creating serverless solutions that go beyond FaaS. We still see people associate FaaS with Serverless but Serverless is way more than Functions.
   
   I actually came across this article in The New Stack that claims that serverless is failing AI Agents. The problem is that they are associating “serverless” with “functions”. I plan on doing a more detailed analysis of this article soon but I think it shows that people are still misunderstanding serverless within the industry and are creating artificial limitations within their mind due to this misunderstanding

3. Growth in Serverless data services. We have seen services such Redpanda and Databricks grow showing that Serverless streaming as well as Serverless databases and data lakes are popular. I think we will see growth in this area to complement the growth in Agents and AI in general.

But enough about the future, let’s see what happened in the past. So let’s jump into the January 2026 serverless news...

Serverless Industry

• Can Serverless 2.0 Transform How Apps Scale in 2026? (Yes it can)

• How Serverless Architecture saved a $2.3M Deal.

• Why Serverless Risk Demands Identity-Aware Security

• Kraken uses Serverless Architecture to keep their Grid Green

Startups

• Runpod hits $120M ARR, and it began with a Reddit Post

• Cloudflare acquires Astro, which may simplify webdev on Cloudflare

Product Launches

• Microsoft releases MCP Server Support for Azure Functions

• Elastic’s Serverless Cloud is now GA on Google Cloud Marketplace

• Dev.co introduces full-stack Netlify Deployment for Jamstack

Runtimes

• WASM still Expanding Frontend Uses 10 Years Later

• What’s next for Azure Containers?

Data

• Databricks Highlights Lakebase Demo which Integrates Serverless Postgres

• Machina Sports turns Real-Time Sports Data into Engaging Fan Experiences

AI

• Luchen Technology launches a serverless fine-tuning SDK (and it costs $1.19)

• Seeweb launched a serverless GPU to overcome GPU shortages

• Cloudflare acquires Human Native, an AI data marketplace

Learning

• How to Use Serverless Computing for Your Cloud Projects from Analytics Insight

• How to build a database without a server

• Build a serverless, post-quantum Matrix homesever

Closing Thoughts

So this month, I saw more news than maybe the past few months. A lot of it was related to acquisitions and Cloudflare has been really busy. Elastic has also been busy with GCP and AWS. They are also making major AI investments to expand their offering.

Runpod is a neocloud that is growing and we saw a few others launch or expand recently. We also see companies more openly talk about their serverless journey.

All in all, I think this is great for the market. While I would be remiss to state that serverless is the what all compute should be, I think the cloud should take a “serverless first” approach and then move to other runtimes or platforms if that doesn’t work for the use case. These market trends, especially in the age of agentic AI and MCP servers, show that not only is serverless growing, but many organizations are seriously considering a “serverless first” approach.

As promised, I will try to cover more stories in depth in my newsletter for subscribers to dive deeper than this monthly recap. I wasn’t able to get to it this January but February on, I will try to do at least once every other week and then increase frequency.

See you all in February 2026!

—Photo courtesy cottonbro studio on Pexels—

I personally took a trip to the Pacific Northwest to meet my partner’s friends and family and I will say, it was a fun time! Now, I am embarking on my new New Year’s Tradition, skiing in the New Year in Tahoe!

This year has been a big year in serverless. If I had to sum up the biggest tech trend of 2025 into one word it would be AGENTS. Agentic AI has really been blowing up this year.

A quick summary on agents; they are essentially smart AI applications that can act autonomously. A great real world example would be an agent that summarizes your daily emails into a document and then books meetings onto your calendar. McKinsey has a decent write-up here if you want to do a deeper dive.

To me, serverless will be the biggest winner in this new Agentic AI world. I think we will see an even larger explosion of Agents in 2026 but that code will need to live somewhere and serverless compute will be perfect for it. After all, why worry about infrastructure when you are trying to rapidly deploy agents?

The market is already showing this. Vercel launched Fluid Compute and Inference just raised money this year as well. There are several other examples in previous newsletters so give them a read. The future is looking bright for serverless.

So let’s jump into the December 2025 serverless news...

Serverless Industry

• Nscale acquires Future-Tech to accelerate global datacenter delivery

• Selective Data Protection against Memory Leakage Attacks on Serverless Platforms

• Scientific Papers on improving Performance Efficiency of Serverless Platforms

Startups

• Akamai Technologies acquires WASM company Fermyon

Product Launches

• AWS Announced AWS Lambda Managed Instances

• Cloud Run continuos deploy is in Public Preview

Runtimes

• Render vs Vercel, an interesting take

Data

• AWS EMR Serverless eliminates local storage

• IBM Acquiring Confluent, a Kafka streaming company with serverless options

AI

• Vast.ai launched serverless orchestration for GPU Cloud

Learning

• Build and Run Secure, Data-Driven AI Agents

• FinOps and Serverless

• Architecting Planet Scale, Modern Apps in the Cloud (serverless is here)

• AWS Blog - Migrate MLflow tracking servers to Amazon SageMaker AI with serverless MLflow

Closing Thoughts

This month (and year) has been a big one for us. Not just AI Agents but also other areas in serverless such as databases and compute. Cloudflare, for example, made several improvements to their cold start time and Google Cloud Run GPUs are now GA.

There are also many other stories of this year that show the explosion of AI use-cases in serverless as well as new players entering the field. We also ssaw improvements in serverless eventing and serverless databases. It’s pretty amazing to see serverless evolve from functions to being an important piece of cloud computing.

Looking towards 2026

I hinted at major changes for 2026 and here is what I think I will do starting in 2026.

• I will continue to do the monthly newsletters that summarize the serverless news with links to follow. I will also provide commentary on that month. These will launch the first week of the month, covering the previous month.

• Substack subscribers will see more short form newsletters where I cover a specific serverless topic or provide commentary on something in the news that I think warrants it’s own newsletter. The goal is to give you more valuable information but also make it “bite-sized” so tha you can read it on a break.

• I did attempt to launch a podcast in 2025 but due to a variety of events, I was unable to do so, so I will try that again this year, likely in Q2.

• I do have a YouTube Channel and may experiment with shorts in late Q1 and see how you all react to that.

Let me know your thoughts and I can’t wait to see you all in 2026

See you all in 2026!

—Photo courtesy George Dolgikh on Pexels—

I always found it weird that we practice thankfulness once a year. Really, it should be an “always thing”. You know, we all have our ups and downs and sometimes life seems cruel but we also have things to be thankful about. It may not something tangible like a mansion or whatever but things like health, loved ones, etc.

Anyway, KubeCon North America was a few weeks ago but sadly I did not make it. I sustained a back injury that pretty much took me out of commission for nearly two weeks. I am much better now but needless to say, I was bummed. The Cloud Native community is one of my favorite communities to vibe with.

There is a YouTube Playlist though so I encourage everyone check this out.

Looking at the event, there wasn’t a ton of Serverless talk but maybe we can change that for next year. After all, serverless and Kubernetes are not mutually exclusive. I am working on a blog post that I should drop in December that explains this.

While we are talking about Serverless, let’s jump into the news of November.

Serverless Industry

• The serverless security market is rapidly growing and expected to hit $25.3B by 2032

• How one financial firm processed $2B in fraud claims with serverless.

• Why Tax technology needs serverless.

• Serverless computing is one of the top trends in Cloud Computing for Businesses

Startups

• Cloudflare to acquire Replicate.

Product Launches

• NVIDIA Cloud Functions now a part of DGX Cloud

• Grafana and GitLab team up to bring serverless CI/CD observability

• Traefik Labs launches a new application intelligence platform that can help you with serverless

• Fermyon WASM Functions is GA on Akamai

• Aviatrix launches zero trust for cloud native and AI workloads (including Serverless)

Runtimes

• Cloudflare Introduces Remote Bindings for Local Development

Data

• SingleStore unveils AI focused enhancements for real-time data and serverless workloads!

• Iceberg REST Catalog now supported in Google BigLake metastore (serverless data)

• AWS introduces Amazon MWAA Serverless

AI

• Why your Enterprise Multi-Agent platform should be serverless!

• User Aware AI Agents and MCP servers just makes sense with serverless.

Learning

• Build and Run Secure, Data-Driven AI Agents

Closing Thoughts

Cloudflare was in the news a lot this month. Their acquisitons and new offerings had to be acknowledged. Of course they are one of the early serverless companies so they will show up a lot.

We had way more serverless product launches this month than in prior months which I think is great. Rewind the clock maybe 6+ years and people were not talking about serverless offerings in this way.

Of course, so were AI Agents. Honestly, I think 2026 is going to be the year of agents which will be great for serverless. 2025 was really more of us just kind of figuring out agents but now we are going to see some serious workloads.

I will be changing the newsletter format so look for another entry sometime in mid-December to explain that format in more detail. The TL;DR is that I just want to publish more so I will be releasing a few different types of newsletters so stay tuned!

See you all in 2026!

—Photo courtesy Pixabay on Pexels—

I did get a lot of game time with “Ghost of Yōtei” and I also went to Denver to catch-up with some old friends. I also played a bit with different serverless technologies.

I am definitely that person who gets bored at times then decides to play with a new technology and then just moves onto a new thing. Every so often though, there is a technology that really sticks with me. Knative is one of those.

In case you are unfamiliar, it is a piece of open source software that enables one to turn their Kubernetes cluster into a serverless platform. Well, this month, it officially graduated at CNCF! This is basically the CNCF “rubber stamping” the project and declaring it fully mature.

I am a huge fan of the platform. You can read my old Serverless Eventing blog to see that. But this graduation means so many different things. One thing I am looking forward to seeing is people using it for AI Applications.

I plan on experimenting and write more about that on my blog and here but for now, let’s jump into the news.

Serverless Industry

• Serverless Computing market expecting to see a big move in 2034.

• 7EDGE cuts cloud costs thanks to serverless

• Serverless is a key tech powering the cloud

• Cloud functions market hits a new high

Startups

• Crusoe raises $1.38B to make AI data centers (and offers serverless GPUs)

Product Launches

• Coreweave launches serverless reinforcement learning

• Tria Federal Unveils Tria Forge. A Serverless AI Platform geared toward security and “codeless AI” development

• IBM introduces serverless fleets with GPUs for high-performance

Runtimes

• Your next app may run on “paranoid lambdas”

• Cloudflare achieves a 99.99% warm start rate!

Data

• How are serverless databases changing how we see data?

• MariaDB introduces a unified database platform (that has serverless functionality) to accelerate AI application development

• DeltaStream Unveils Next-Gen fusion platform for real-time analytics

AI

• Yotta launches Shakti Studio, a Serverless AI Cloud meant to speed up Enterprise go-to-market.

• Amazon Bedrock simplifies access with automatic enablement of serverless models.

Learning

• Beat the Java slow startup time on serverless

• Build a serverless API with Bun and Hono

Closing Thoughts

Serverless clearly is expanding WAY beyond the definition of runtime functions. We are now seeing serverless databases and AI. In 2026, I predict we will see even more serverless AI in the form of serverless MCP servers and Agents. We are already seeing that but I expect to see some real innovation next year.

November is also KubeCon + CloudNativeCon - North America 2025. I am always excited to go to this event. I learn so much and there are always great talks. I want to learn more about the serverless stuff like Knative, KEDA and WASM.

I will be sure to report more on what happens in serverless in the next newsletter.

See you all in November!

—Photo courtesy olia danilevichon Pexels—

I personally had a great September. This year, I had the privilege of being able to help organize the first ever San Francisco Bay Area Kubernetes Community Day (KCD). These are volunteer organized events that are sanctioned by the CNCF.

I was privileged to work with amazing people like Rey Lejano, Matthew Cascio, Natalie Lunbeck, and Lisa-Marie Namphy as we create a day long event with amazing speakers and attendees.

While there wasn’t a lot of talk specifically around serverless runtime, most everything mentioned was relevant to a serverless platform as well as a Kubernetes based platform.

I would encourage everyone to look for a KCD near them next year to participate in. Open source is only amazing because of the people who contribute to it.

On that note, let’s jump into some of the serverless news you may have missed.

Serverless Industry

• Serverless Computing market expecting to see a big move in 2033.

• Moody’s reduced change ticket turnaround by 83% by adopting a serverless architecture

• How serverless is transforming retail tech in a multi-cloud world

Startups

• Guepard Secure $1.2M to launch serverless data platform for AI Agents

• Trigger.dev launches to build and deploy serverless (fully-managed) AI agents and Workflows

Product Launches

• Google Cloud Run GPUs are now in private preview for Worker Pools

• Tria Federal Unveils Tria Forge. A Serverless AI Platform geared toward security and “codeless AI” development

• AWS integrates LocalStack with VS-Code Toolkit. The goal is to streamline serverless development

Runtimes

• Cloudflare brining Node.js HTTP servers to Cloudflare Workers

Data

• Mother Duck announces European Cloud Region to bring Serverless Analytics to the EU

• Enterprise data teams cut operational costs by 40% (and serverless helped!)

AI

• AWS Hires two VPs to drive Agentic AI for Agentcore and Kiro

• Is Cloudflare’s Edge Computing Momentum helping in an AI world?

• Modal Labs raises $80M to simplify Cloud AI (with Serverless)

Learning

• Building Resilient Multi-Regional Cloud Run Services

• How to Start Your DevOps career (spoiler, learn serverless)

• Build a full fledged CMS on Cloudflare Workers

Closing Thoughts

This month wasn’t as heavy in the news but I do think we are seeing some trends. One, people are saving money by moving from traditional architectures to ones that leverage more serverless. Two, Serverless as an industry is growing and is expected to hit $9B in value by 2033. Three, AI loves serverless. AI is supposed to make development easier. Serverless makes development easier. It only makes sense that they work together.

I love seeing serverless startups continue to take off and established serverless companies expand. I also love seeing traditional companies adopt serverless either internally or as a product offering.

Let’s see what October brings!

—Photo courtesy Tomas Anunziata on Pexels—

This will be the first go of our new format. I will break down the news into 6 categories, technically eight if you include the intro and outro. First will be “Serverless Industry” which will cover serverless as an industry and general serverless news. The others are pretty self explanatory. So let’s jump in!

Serverless Industry

• CXO Today talks about Serverless being the next frontier for “cost-effective cloud”. Makes sense, pay only for what you use without losing functionality!

Startups

• How Startups can optimize cloud costs (with serverless).

Product Launches

• Elastic launches Serverless Log Analytics. This will be a lower price than similar offerings from Elastic.

• Oracle Launches Exadata Service for AI, Compliance, and Location-Critical Workloads, a fully managed serverless database.

• 7EDGE Unveils API Gateway Expertise to Accelerate Serverless Innovation

Runtimes

• AWS Lambda integrated with GitHub Actions to Automate Serverless Deployments

• Serverless Containers are a growing trend!

Data

• MariaDB buys its former subsidiary SkySQL to enhance its agentic AI and serverless capabilities

AI

• NextGen Cloud launches on-demand access to Open AI GPT-OSS

• Cloudflare targets FedRAMP High by 2026.

Learning

• A blog on Growth Jockey talks about how serverless architecture can enhance serverless websites.

Closing Thoughts

I think the big theme that we are seeing here is that serverless is still growing and enterprises are seeing it as an actual product that they can use versus just being glue to connect things.

It is also great at cost savings. I know I am beating a dead horse here but serverless allows you to pay only for what you use. Granted, you can let costs go out of control if you don’t set proper guardrails. But if you do plan, you can see amazing savings with serverless.

We also see companies ramping up their serverless AI offerings. Cloudflare is seeking FedRAMP High for their serverless AI products and MariaDB re-acquired SkySQL.

Then of course we have the tools for developers such as serverless logs and an API gateway.

Let me know what you think of the new format. I think I will experiment with it a little bit until the end of 2025 and we’ll see what we get!

—Photo courtesy Luana Scorsoni on Pexels—

Kids are out of school, people are on vacation, the grills are being brought out for cookouts. Yes, summer is here. However, just because you are taking a break doesn’t mean that the serverless world did.

I certainly did my serverless reading and I will use this newsletter to tell you what I learned. Also, at the end, I will tell you about the new newsletter format starting in August 2025.

Building a Serverless Environment

Often times when I talk about serverless I talk about deploying application on a serverless runtime or using a serverless database. These are, of course, important things but what about the actual build process?

I read DevOps.com and I came across this article on Serverless CI/CD. Now to level set on CI/CD. CI/CD stands for “Continuous Integration/Continuous Delivery”. It is the practice in which you are able to build your application, deploy it, monitor it, and iterate on it. Octopus Deploy actually has a very good article on the concept and it includes the famous DevOps loop (see below).

So what does this have to do with serverless? Well, I think with the ease of serverless, many people do a simple “push to deploy” with their code from either their git repository or even from their local development environment. I mean it’s easy right? Package the code and push it, it’s live!

For personal projects, that is probably fine. However, when you are working as a team on a large open source project or an enterprise application, you need to put in guardrails to make sure that nothing goes awry and there is a good practice to maintain your release strategy.

A good CI/CD strategy will do that. It can help improve code quality, simplify rollbacks, reduce costs, and more. I am not here to tell you which CI/CD DevOps tool is the best. I personally am partial to GitLab but there are many options out there. Find the one that works best for you.

The Serverless AI Startups Keep Coming

This startup showed up in my feed a lot this month. South African startup Cerebrium raised $8.5M from Gradient. What does Cerebrium do? Well they do serverless AI. That is, they are a serverless platform for hosting AI applications such as chatbots, voice assistants, and whatnot. It scales up and down as needed.

We have seen several startups like this lately, all trying to be serverless. I have said it once, I will say it again. When it comes to AI Applications, serverless makes the most sense. Pay only for what you us and don’t worry about infrastructure. No developer cares about infrastructure. Write code, deploy it, and trust that your platform can handle the scaling.

Another startup, Blaxel, just raised $7.3M to be the “AWS for AI Agents”. They are trying to be a cloud for AI Agents. They can handle serverless APIs, MCP Servers, and more. They are trying to fill in a gap that, quite frankly, the larger cloud providers are struggling to do now. Again, let’s make deploying agents and MCP servers easier. Serverless!

You can see a neat interview with the co-founder, Charles Drappier, here on Pulse 2.0.

Serverless inferencing is also becoming a bigger deal so I predict more and more of these stories in the future which I will cover in the new format (more on that later).

More WASM!

Microsoft is doubling down on their Micro-VM designed for WASM workloads. This tool is called Hyperlight. Hyperlight WASM is a Rust-based web server that can execute multiple functions on micro VMs. It can support other languages, don’t worry.

I still wonder what the future of WASM will be. WASM is very promising and major companies like Microsoft are doubling down on it. But it does seem to be lagging when it comes to AI workloads. I think that fact is what’s causing people to not talk about it as much. However, not every application is an AI application. We still have non-AI apps out there. WASM may very well be a great use case for that.

Serverless Remote MCP

Cloudflare has launched a Remote MCP Server! Cloudflare is pretty much synonymous with serverless nowadays, at least when it comes to runtimes. Well now they offer a remote MCP server option for you to build and host with no infrastructure. Cloudflare is doing some really cool things and I look forward to seeing more of what happens here.

New Format

I will be exploring a new format starting with the next edition of the newsletter. I will be categorizing stories to cover Startups, Product Launches, Runtimes, and Data. This should make the newsletter easier to read but also deliver more information.

Admittedly, my previous formatting was a bit haphazard which would make the stories harder to read (and harder to write). By categorizing the information and including smaller blurbs, I can deliver more content. I will also include an op-ed section of sorts where I just kind of brain dump my overall thoughts on serverless trends.

Closing Thoughts

Serverless is here to stay. We are seeing more and more serverless startups getting funding. We are seeing major players invest in serverless options, and we are starting to see people think about how serverless will fit into their processes and platforms.

What will the serverless world bring us next? Well stay tuned to find out!

—Photo courtesy Anna-Louise on Pexels—

This month, I read a lot of stuff similar to what you have read in the past. A lot of reports suggesting that serverless cloud is growing and expected to hit unprecedented numbers by 2030.

This month was a bit lighter on the news side but I did want to take the time to write you all.

A Decade of Serverless

AWS opened the floodgates for serverless with the creation of Lambda in November 2014 (GAing in April 2015). The creation of this Functions-as-a-Service (FaaS) framework resulted in the first “serverless” platform. Now technically, one could argue that PaaS was serverless too but most PaaS platforms had some form of infrastructure that you had to manage.

According to Technology Magazine, the FaaS industry is a $30bn global market now. This shouldn’t come as a surprise as many companies are trying to find ways to reduce their overhead. Of course, just like with anything, there are pros and cons to functions.

Many of the cons such as vendor lock-in can be attributed to the opinionated natures of FaaS runtimes. But with containers, you can potentially bypass this. This could be why Cloudflare launched containers in beta this month.

Serverless continues to grow and innovate such as the shift from functions to containers and WASM in serverless circles. Functions will continue to exist but having more options will reduce friction seen in serverless.

Even heavily regulated industries like healthcare are starting to see the truth of serverless.

As a side note, I found an interesteing article where someone built an app for under $10 a month with serverless.

A Serverless Data Lake?

Yes. Now that Databricks has acquired Neon they are looking to introduce a serverless data lake called Lakebase. Now admittedly, data storage and analytics aren’t my strong suit so I had to dig into this.

To level-set, a data lake is essentially a massive database for storing data in it’s raw format. Once the data is collected and stored, data scientists and analysts can then clean the data via an ETL pipeline or something similar so that they can try to gather valuable insights.

With Neon and their serverless Postgres offering, Databricks is looking to offer a data lake that is serverless in nature. That is, the storage of the data and the processing of the data are separate.

Per Databricks, this will allow businesses to keep their data with their analytics platform. This shows how serverless continues to innovate and how it is WAY more than just functions.

If you want to learn more, you can review this blog post!

Cloud Run GPUs @ Google Cloud get a few Upgrades!

Google Cloud Run GPUs are now GA for Services! This is true for the NVIDIA L4 model of GPU. But also…. GPUs for Cloud Run Jobs are in Public Preview (Google Cloud terms for “beta”). What does this mean for serverless an AI? Well with services (Cloud Run instances needing an HTTP/S endpoint) you can create endpoints using tools like Ollama or vLLM in a container to inference with LLMs such as Gemma or DeepSeek. These GPUs are on demand and will only run when a service is running (e.g. someone hits the endpoint).

Jobs are used for batch processes usually. So say you wanted to use an LLM to read documents in a storage bucket and store the results in a database. A Cloud Run Job would be perfect for this use case. Now it can use GPUs as well!

Closing Thoughts

Serverless is the best platform to start building applications in the cloud. There will always be a need for VMs, container orchestrators like Kubernetes and other application platforms but serverless should always be the starting point.

We are seeing major companies doing big things with serverless and helping us evolve past the notion of simple functions. We see serverless GPUs, Serverless datalakes and more.

I will leave you with this article from InfoWorld that talks about many misconceptions around serverless and how serverless is going to be an incredibly powerful tool in the age of AI.

Until next time!

Update

In order to provide more value to my readers, I am going to be making some format changes that will be visible in August 2025. You may see some samples in the July edition. I haven’t decided exactly what I want to do yet but I will make a formal announcement of the formal changes and release schedule in the July 2025 edition.

—Photo courtesy Nitin Dhumal on Pexels—

I managed to get a few ski days in before the season officially ended and we had to move out of our ski lease. Now comes the time where I put my ski gear away until December. But hey, serverless news is always fresh (admittedly not my best transition)!

I opted to use a photo of my dog getting caught in the snow (don’t worry, she was fine) to signify the serverless storm. This month I read a lot on how many cloud native application businesses are moving to serverless and a new era for enterprise efficiency being lead by AI and serverless. This seems to indicate that AI is going to bring in a serverless storm so let’s talk about it.

AI as a {Serverless} Service?

Rafay has released a Serverless Inference for NVIDIA Cloud Partners. The idea here is that it will allow these NVIDIA Cloud Partners to improve their margins on GPUs. This will allow the providers to no longer focus on just being “GPU-as-a-Service” but “AI-as-a-Service” (I have called it Inference-as-a-Service in the past).

What does this mean? Well today cloud providers are offering GPUs for their customers to use to either train or serve their models. Nothing wrong with that setup but the margins on hardware are pretty low. With the new Rafay software, the providers can now better monetize the act of inferencing itself.

One way to think of it is like going to a restaurant. I say that I want a burger. With the GPU-as-a-Service model, the restaurant gives me the hamburger meat, lettuce, tomato, pickles and bun and then a table to assemble and stove to cook.

AI(Inference)-as-a-service is the restaurant giving me a ready made hamburger where I can charge extra due to labor, atmosphere, whatever. As the end user, I get what I wanted, food.

Sounds a lot like serverless right? Instead of me having to create my own server with inferencing software and a GPU attached to inference with my model, I can just use the inferencing software that my cloud provider gives me.

I predicted that inference-as-a-service was here to stay and this move by Rafay and NVIDIA are showing how this is happening.

Rafay had a great presentation on this concept if you want to dig in further.

Rise of Serverless MCP

In April I talked about the rise of the Model Context Protocol (MCP). Cloudflare has announced an MCP Server but so have many others!

A part of Google I/O, Google announced Cloud Run MCP Servers and an AI Studio integration. Cloud Run is Google’s serverless platform that relies on containers for their runtime.

AWS also announced their own Serverless MCP server. This of course integrates with AWS Lambda. While I didn’t find a specific product from Azure, they do have a blog post that shows how to run MCP servers on Azure Functions. It’s like a DIY kit for MCP Servers.

The fact that “the big three” are validating MCP servers can signal that this is a concept that is here to stay… for now. AI is a space that is constantly changing but I think that MCP will evolve into something else rather than go away completely.

The other interesting fact is that they are focusing on the serverless use-case. Using MCP should be easy. Making someone spin up a Kubernetes cluster or a VM to deploy an MCP server seems a bit backwards in 2025. Of cousre you want to make that runtime easier to deploy and that’s why serverless is incredibly common for MCP Servers by the cloud hyperscalers.

More on Serverless Startups

In the past I have talked about Neon, the serverless database often used for RAG implementation. Well DataBricks is buying them for $1B. DataBricks has become one of the leaders in AI and Data and could be seen as a late stage startup by many. They haven’t been acquired or filed an IPO but my guess is that they will within 3 years.

DataBricks themselves are a serverless company in the sense that they have some first-party serverless offerings. Buying Neon would ultimately open up new options for their platform. This will help them have native RAG options on their platform allowing their customers to have a more customized experience. They could also offer it as a service to the hyperscalers.

Another startup, HostColor.com, has created a serverless Wordpress offering. Wordpress is an open source Content Management System (CMS) and is said to power roughly 40% of the internet. My site, jasonsmith.io, was a Wordpress site up until a year and a half ago ( read more about my migration here ). Great product but more than what I needed.

Anyway, I thought this was pretty cool as most deployments of WordPress that I have seen were monolithic. While there are many WordPress hosts out there, what I am curious to see is if this helps them provide better performance or an overall better offering.

They appear to be using a FaaS based solution so I am curious what that looks like from an architecture standpoint. My inclination is that the decomposed all of the application functions into actual function-based microserver.

If you want to see a serverless container-based solution, feel free to check out my demo here.

Guarding of Serverless

As more and more people use serverless technology, securing that technology will become important. Now there are two personas to concern here. There is the serverless provider and then the serverless user. Ideally, with serverless, the user has to consider less about security, at least when it comes to infrastructure, but they are still responsible for their applications

Serverless runtimes require people to rethink Application Security (AppSec). I found a neat article on HelpNetSecurity.com that talks a little about the considerations.

I want to encourage people to read this article on their own but I wanted to highlight something interesting that was said.

“With serverless, you’re no longer dealing with long-lived servers or predictable runtime behavior. Instead, you’re working with short-lived functions triggered by events, often running in unpredictable sequences. This makes it much harder to apply traditional security tools that expect stable infrastructure. What’s needed instead is a dynamic approach, one that understands how these functions behave in real time, within the context of the application’s actual workflows.”

Now they specifically called out functions but the same rules would apply to serverless containers and WASM. Security for serverless needs to be thought about in a different way.

And just because you don’t have access to the infrastucture doesn’t mean that you are hack proof. After all, there was recently an attack that targeted serverless platform. While I don’t have specific news about a new serverless security platform or tool, I want to remind people to always take their security seriously when on the cloud.

Closing Thoughts

Serverless is growing, this is a fact. I believe that this AI revolution has been something of a tipping point for the serverless movement. It is now reshaping FinOps as compute is becoming even more “pay-as-you-go” than it has been in previous years.

This past month we saw a major acquisition of a serverless firm, someone taking 20-year old software and making it serverless and serverless MCP servers being provided by major clouds. A lot of this, again, is being driven by the rise of generative AI.

I look forward to seeing what we have next month.

By the way, if you want a crash course on serverless AI/ML, check out this post from KD Nuggets.

There were a number of announcements that you can see here but that’s not what you are here for. You are here to learn more about serverless computing and what happened in April.

Reading the news this month, I saw a number of articles on AI (of course), Wasm, and a few serverless startups. I also saw a few well established companies with stock symbols invest heavily in serverless. We even had a big announcement regarding serverless containers!

This excites me as it shows that serverless is getting its flowers ( receiving it’s deserved attention ). As I have stated before, I have been a major fan of serverless for years. The term serverless didn’t take off until circa 2015 with the launch of AWS Lambda. Seeing serverless being taken more seriously as a technology makes me happy.

But let’s jump right into the news.

Cloudflare meets… Containers?

I have spoken at length about serverless containers. It is undeniable that the container is the basic compute unit for the cloud. Yes, VMs still exist (and will continue to exist) and WASM is becoming a promising technology but containers are the core of cloud native technology.

Now again, when many people talk about serverless, they are thinking about functions. Obviously, serverless is more than functions. Well imagine my surprise when I learned that Cloudflare felt the same way.

Yes, containers are coming to Cloudflare. Cloudflare Workers, which has historically only supported a FaaS-style architecture, will be supporting containers in June of 2025.

Cloudflare has never really ventured much out of the FaaS world. It is amazing that they are embracing containers. They are recognizing the portability of containers. By supporting containers, they are entering the “cloud native” world and joining the “serverless container” world.

It will likely open a lot of opportunity too. We may see serverless GPUs evolve there, for example.

What the WASM?

WebAssembly (WASM) promises to be the next step in cloud binary packaging. Fermyon is positioned to be THE WASM company right now. I got to see them at our Google Cloud Next 2025 Day Zero event, Containers @ Next. They displayed something really cool on GKE.

They showcased the Spin Framework. This was created by Femryon but is now a part of the CNCF. It is a framework to deploy and manage WASM apps and can be run on Kubernetes.

One of the things that has been holding WASM back is that it lacked a reliable autoscaler. It is kind of like what we saw with Docker in late 2013. People saw the containers and thought that they were great but what happens when you need to run 100 of them?

Spin Framework on Kubernetes can help serverless WASM expand into new territories. Kubernetes already has great market share and the CNCF has started to explore WASM projects, it only makes sense to use Kubernetes as the platform for WASM.

Microsoft recently created a microVM for WASM workloads. This is a somewhat different approach from using Kubernetes but I think we need multiple ideas.

Serverless Joins the MCP

In case you haven’t been checking the tech news of the past few weeks, Model Context Protocol (MCP) has been the talk of the tech town. A quick summary, it is a protocol that allows developers to communicate with agents more easily. It was created by Anthropic back in November of 2024 but it has really taken off over the past month or so.

This graphic is a great demonstration of how this is supposed to work.

Before MCP, all of your agentic applications would have to make individual API calls to your LLM. MCP operates in a client-server model so your apps just need to connect to the MCP service and that server connects to the LLM.

Naturally, the server needs to live somewhere. It’s not a server in the traditional sense either. In other words, it doesn’t need to be an actual piece of hardware or even a VM.

Our friends at Cloudflare announced a serverless MCP server. I think as time progresses, we will hear more about serverless MCP and I will gladly report it.

As “Inference-as-a-service” startup!

Let’s do a speedrun of startups

Inferless, a serverless GPU company has launched. Read the interview with the CEO here.

NScale launches a serverless AI Platform for inferencing. It is based in Europe so this serverless AI concept is going global.

OVHCloud launched a serverless AI endpoint service.

Closing Thoughts

Serverless is taking over. I was pleased to read this post in InfoQ that talked about work my buddy did when he was at Capital One. They are ALL IN on Serverless. Banks tend to be very hesitant when it come to adopting the cloud and not only did Capital One go all in on the cloud, but they went all in on serverless.

I think we are going to see more and more serverless. I think MCP and even Google’s own A2A protocol will result in a rise of more serverless GenAI use cases.

SaaS companies are already evolving from monolithic to serverless architectures. As they begin to adopt agents, I fully expect for them to adopt serverless at a faster rate. We are hearing people clamour for more agentic AI and LLM use cases. DeepSeek runs amazing on serverless platforms as do other LLMs.

I fully anticipate more startups coming into existence dedicated to serverless AI. As more and more happens, I will gladly report it here!

—Photo courtesy … Me—

Anyway, you didn’t come here to listen to me talk about skiing and snow. You came to learn about serverless!

There have been a number of interesting stories as of late.

Agentic AI and Serverless

I have spoken often about Serverless AI and Inference as a Service. I even published a blog post with Google on inference-as-a-service last month. One thing that I often state is that serverless is the perfect platform for generative AI. Not so much the training (though who knows what the future holds) but definitely with regards to serving/inferencing.

As Generative AI has matured, we need to look past LLMs and look more towards agents. At least this is what Marc Benioff said in a recent interview. Generative AI is going to be less about building “bigger and badder” LLMs and more about building software that can leverage LLMs to perform tasks with little to no human input.

Needless to say, the best way to do this would be with serverless runtimes. Let the developers build agents with the least friction possible. We are seeing a rise of serverless runtimes built for AI. Even NVIDIA launched their DGX Serverless Interface recently. They see the value in serverless for AI.

We need to think beyond simply building chatbots but building agents capable of doing more than simply chatting. Agents need to be able to scale up and scale down as needed. This is because these will arguably be very “bursty” workloads so maintaining infrastructure just doesn’t make sense.

Some Raises, Some Developments, All Serverless

Featherless.ai also just raised money from Airbus Ventures. They focus mainly on serverless inferencing for AI. This space is truly growing as developers are wanting to find simple ways to leverage AI without having to take on unnecessary technical debt.

In the past I talked about Fermyon on their investment in WASM as a runtime. Well just recently they teamed up with Akamai to bring serverless AI to the edge. (If you want to learn more about Edge Computing and Serverless, feel free to review this IEEE article. )

You could, in theory, have mobile applications or kiosk application that could connect to the internet leverage AI runtimes on a CDN. That’s pretty wild when you think about that. I mean, Cloudflare has been doing similar things with their functions and their global CDN but this feels a bit different. For one, it’s WASM which is proving to be a very strong runtime. I haven’t heard of someone bringing WASM to the edge before so this can really change the game. If WASM is as fast as they claim, this could not only address cold-start times but given faster compute experience.

This article talks about serverless payment gateways and honestly, serverless-edge may be the best way to execute!

Vercel, a truly serverless company, has now announced Vercel Fluid. In short, it’s a way to allow serverless functions to behave more like a VM. You still get the elasticity of serverless functions but you are handle multiple requests. This can also help address the cold start issues that plague FaaS. Now I am a tad skepitcal of the claims but I am very interested to see where this goes. I don’t know that it will be as scalable as pure functions but I am also happy to be proven wrong.

Journey to Serverless

In Forbes, Kirimgeray Kirimli is the CEO and Cofounder of Snapshot Reviews and Director of Flatiron Software, talked about how he moved Flatiron Software to serverless.

He does go over a few of the pain points and “triall and error” that they engaged in to get where they were going. However, this story is a great one as they talk about how the simplicity of serverless overall made them more productive. I encourage you to read this on your own but let me leave you with a quote…

“At the end of the day, infrastructure should work so seamlessly that you almost forget it’s there. And that’s the beauty of serverless—it clears the path for building what’s next, letting us focus on solutions that drive impact rather than roadblocks. As the technology continues to evolve, I’m confident serverless will play an even bigger role in shaping the future of how we design and maintain systems.”

Closing Thoughts

Serverless is PERFECT for Agentic AI. “SaaS is dead, long live Agentic AI” is what I have been hearing lately. While I think that may be a bit hyperbolic, there is a lot of truth to that.

People will begin interfacing more with agents vs. click-ops through a website to get the information that they need. It’s possible that the web browser, as we know it, will no longer be used. We may have agent portals that exist and you just make the request to an agent and it returns an answer.

At the end of the day, serverless is what will ease the deployment of Agentic Applications. Stop worrying about infra and start worrying about your apps. The major investments into serverless for AI should be proof that this is a viable market!

—Photo courtesy Bich Tran on Pexels—

Serverless is something that I have been a passionate fan of for quite a while. It wasn’t always called serverless but the idea that one could deploy a runtime and not worry about infrastructure was always a win for me.

I used to have a server in my office during the college years. All the fun provisioning I had to do to get a LAMP stack running was always a pain.

But some people tell me that it’s awful. Builtin says that “the hype” is over. But this article makes the major mistake of confusing FaaS with Serverless. We really need people to stop doing this. Say it with me now “Serverless > Functions!”

InfoWorld recently had an article that had a more balanced approach. They mention that serverless isn’t a cure-all and I would generally agree with that. Even I will tell my customers that there are times when Kubernetes makes more sense that a serverless platform like Cloud Run. But they also take the same approach of limiting serverless to FaaS and starting with that premise makes things problematic.

TechBullion recently called out the advancements in serverless and it’s role in the next wave of cloud native innovation. Engineers addressing the cold-start issue, vendor lock-in becoming less of an issue, and more advancements in serverless is showing that it will be the next wave of cloud.

Snowflake recently talked about their serverless platform. For those who aren’t familiar, Snowflake is a cloud data storage company that now has ventured into AI. This isn’t purely a compute company, but a LARGE data company. This is less than functions!

CIO Magazine demonstrates how serverless can help Fintech companies modernize and save money. I encourage you to read the report but some highlights include seeing a cost improvement of about 35% and “time to market got shortened by 15% on average”. Financial industry is historically very slow to update their tech. Seeing this is promising and I would love to see more design patterns like this.

Of course we can’t forget about GenAI. Serverless cloud Koyeb is now offering Tenstorrent accelerators! Their hope is to make “inference-as-a-service” more standard (my wording) by supporting serverless AI applications. They seem to be aiming to take advantage of serverless’s low-latency but let’s not forget the developer platforms.

Is serverless dead? Hell no. Have we outgrown FaaS? Maybe. But serverless grew up too and we need to acknowledge this!

—Photo courtesy Monstera Production on Pexels—

The New Stack recently wrote an article listing trends of 2025 in cloud computing and naturally, the growth of serverless was there! One thing that stood out to me was this statement.

”It is inherently suited to microservices, allowing developers to build applications as a collection of small, independent services, improving agility and flexibility.”

For the past decade, microservices have been an important trend in cloud native computing. While Kubernetes is an amazing orchestrator for this framework, it is “too-damn low level” for developers. Serverless is the next evolution of cloud native microservices.

I think we should talk a little bit about the new stories involving serverless that hit in January of 2025. So let’s get started.

Serverless for Machine Learning.

(Fair warning, this is paywalled) I saw this interesting research paper on ResearchGate.com. In short, they looked at how serverless compute was being used for machine learning inferencing. Remember when I mentioned “inference-as-a-service”? It looks like the idea has caught on with machine learning cirles.

The nice thing here is that we are seeing a new demographic for serverless use-cases. Historically, we have been very developer centric when we talk about serverless platforms. This is showing how versatile serverless can be as it is able to help data scientists be more productive.

We are also seeing the rise of “AI Devs”. The New Stack talks about this company called Modal.com which offers a serverless AI platform. This simplifies development of things such as AI Agents for developers. Simply build, inference, and deploy!

While not AI specific, data processing and ETL (Extract, Transform, Load) are adjacent to machine learning. This article from Analytics Insights discusses how serverless has improved the ETL. Seeing as we are estimated to be have 175 zettabytes of data by this year, it only makes sense that we find a way to scale CPU power as we try to process that data in near real-time.

Remember, one of the major reasons people moved to the cloud in the first place was to turn CapEx into OpEx and make it a bit easier to future proof their technology. Rather than having to buy a bunch of additional servers to offer you the overhead that you need for growth, you just purchase what you need today and often with the “flip of a switch”, get more compute power as needed in the cloud. Serverless gives you that but with way better scalability.

As we move into this next phase of Generative AI, I expect to see way more serverless compute used to help data scientists be more effective. We will end this section by sharing this very awesome HackerNoon article on serverless machine learning.

Further Advancements in Serverless Security

I found this article discussing the growing security trends as it relates to serverless technology, in particular serverless compute. The serverless security market was valued at $2.3B back in 2023 but is expected to hit $25.3B by 2032.

Now I need to put an asterisk on this as the artcle seems to be leaning heavily towards FaaS and as I have always said, FaaS is NOT serverless, it’s just a segment of the serverless platform. Now they do say that FaaS is leading the serverless security market but I anticipate that we will see fast growth in serverless AI, containers, databases, WASM and more.

Many security vendors are now offering serverless security. I mentioned this is a previous post. One example would be my friends at SysDig offers solutions around serverless containers as well as FaaS.

When we do try to focus on cost efficiency, we don’t want to sacrifice security. Cutting corners is never good and as application developers, we have the onus to protect our user data.

A New Challenger Apporaches!

Everyone loves a good “David and Goliath”. We like to see the scrappy character come and take down the giant. While I do not foresee anyone taking down Google Cloud, Amazon Web Services, or Microsoft Azure anytime soon, I think it’s great to have disruptors show up to really shake things up.

I came across a company called Render while reading siliconAngle. They just raised $80M and their goal is to, in essence, revolutionize cloud computing by taking a more “serverless-first” approach. Traditional clouds are complex with all of their offerings and configurations.

From what I can read, their goal is to be “developer-centric” and make it painfully easy for developers and organizations to deploy their workloads onto the cloud. Serverless is how they plan to do it. While it seems like they are focusing on FaaS, my hope is that they disrupt how we all currently think of the cloud and start to change the behaviors of other organizations.

Closing Thoughts

The future is serverless, at least that what CloudFlare said in a recent interview. Why do we, as end-users and develoeprs, need to think about VMs and all the infrastructure that comes with it? For the first 15ish years of the cloud, it made sense, but we need to get past this.

As we continue to try to scale AI and build more agents, we will need to solve for scalability AND cost. Serverless can do this. As we start to adopt more serverless, we will need to consider additional things like “how do we secure serverless” or “how do we setup guardrails to control cost”.

Inevitably, new players and old players will work to solve this problem and more with serverless. When this happens, we all win! So let’s see where serverless takes us next.

—Photo courtesy Boris Pavlikovsky on Pexels—

We probably shouldn’t wait for a holiday to do those things but at least we have that.

Now, onto Serverless talk. I am a little late in posting this but I was enjoying my holiday with family and skiing with friends for the New Years. I hope you all got a break from reality to enjoy yourselves as well.

I want to break this newsletter down into three segments. I want to look at what happened in December 2024, look back at 2024 as a whole, and then look forward to 2025. So let’s jump in!

A Quick Look at December 2024

Back in August, Google launched Cloud Run with GPUs. This is effectively a “serverless GPU” Well it looks like Azure has followed suit with Azure Container Apps with GPUs. If you are not familiar, Azure Container Apps is essentially Microsoft’s take on serverless containers. It runs on top of Kubernetes using open source tools like KEDA and Dapr. It’s a bit different than Cloud Run as Cloud Run runs directly on top of Borg, Google’s proprietary container system but the general idea is the same, Serverless Containers. And now with another company offering serverless containers with GPUs, we will see a great explosions in Serverless AI.

Our friends at Neon Tech have now implemented native integration with Azure. This will most likely spell growth in Serverless AI within Azure. After all, serverless Postgres databases have a great use case for RAG.

While on the topic of serverless databases, Elastic announced General Availability of Elastic Cloud Serverless. Serverless is huge and we are seeing more and more clouds set this up.

2024 Was “The Year of Serverless!”

December had a few good stories. Forrester recently published The Forrester Wave™: Public Cloud Platforms, Q4 2024. I know that this is paywalled but fortunately, they published a summary in a blog post. In the summary, they highlight three 2024 takeaways for the public cloud. The top listed one was the The serverless-first (and often only) public cloud is here. They do mention Kubernetes which, while in and of itself isn’t serverless, is serverless adjacent. I have gone on and on about the concept of serverless containers and Kubernetes is hands down the best platform to build a serverless container platform on.

One interesting callout in this piece was how Chinese cloud providers were starting to invest heavily in serverless. Admittedly, even I kind of missed this one. As an American working with largely American countries, I am a bit ignorant on what Chinese developers want. I think I need to dig in a bit deeper to understand this.

Back to the topic of serverless containers, Forbes recently published an article on running microservices as serverless. This article was written by Matt Butcher, a Co-Founder of Fermyon. While this focuses more on functions, the general architecture can work for most serverless applications, including containers. They even talk about Kubernetes and WASM.

If I could simplify this, I think we saw three things happen in the world of serverless for 2024.

1. Rise of Inference-as-a-Service. People are discovering that serverless workloads are ideal for AI Agents in the age of GenAI. Companies such as Vultr have invested in created serverless offerings in this space.

2. Rise of Serverless Databases. Neon , Pinecone, Nile, and more have all grown. While not a database, we have seen serverless data streaming as well with companies like Redpanda.

3. Serverless being seen as a viable solution for Enterprise applications

Looking Towards 2025

So what do we expect to see in 2025? I wish I had a crystal ball but one thing is for sure, we will see more serverless. After all, market research still shows that serverless is growing faster than ever.

In 2025, it is likely that we will see more around Platform Engineering. In the past 2 years, the term has grown. As people integrate Platform Engineering into their Development Lifecycle, I anticipate that we’ll see people talking more and more about serverless.

Serverless Security will also become big. As a developer you will still need to figure out how to secure your applications. As you won’t have access to the hardware directly, you may find yourself relying more on managed services to secure your applications.

FinOps for serverless will be big as well. Many people do a poor job at handling their cloud bill and it is very easy to allow a serverless application to exceed your budget if you don’t set up proper guardrails. We turn back to Forbes as they discuss best practices in maintaining the bills.

Of course AI will also play big. Expect to see more serverless in the development of AI Agents and growth in the “inference-as-a-service” market.

Closing Thoughts

There is a lot to think about when it comes to serverless. Just 10 years ago, the idea of serverless was still in its infancy. Ironically, Kubernetes also launched 10 years ago.

In many ways, Kubernetes is a platform that enables the development of serverless platforms. While it isn’t the best platform abstraction for developers, it provides the primitives in order for developers to build their platform.

I think 2025 will be a promising year for serverless given the trend. I see a lot of growth happening with AI Agents and with platform engineering growing as a concept, I could see more platform engineers taking advantage of serverless offerings to empower their developers.

—Photo courtesy Ioana Motoc on Pexels—

Anyway, a lot of serverless news has happened lately. We had KubeCon which I will dive a bit deeper into which really invigorated me. If I had to go to only one tech conference a year, I would 100% choose KubeCon.

I love the energy that we get from KubeCon and I think it’s because it’s an open source conference. At its core, open source is about community. The community coming together to build cool products and share knowledge. It attracts a lot of tech practioners doing amazing things which makes my heart leap!

I love containers and this is what KubeCon is all about! And hey, it’s about serverless too (read on later). I even got to do a lightning talk about Serverless Containers. But then my laptop decided that it didn’t want to work during the demo so whatever. It happens to the best of us.

Outside of KubeCon, there were some other pieces of serverless news so let’s dive in.

Things I Learned at KubeCon North American 2024

There are going to be hundreds of newsletters focusing on the aspects of Kubernetes and AI following the most recent KubeCon so I will let those authors focus on that.

Seeing that this is a serverless newsletter, let’s focus on serverless and how it showed up at KubeCon. After all, per InfoWorld, serverless is making a comeback!

First, to level-set, KubeCon (CloudNativeCon) is an annual conference that usually does one conference in EMEA in the spring and in North America in the fall. This year actually added India and next year will add Japan. The purpose is to celebrate not just Kubernetes but the entire Cloud Native landscape as a part of the Cloud Native Computing Foundation (CNCF). It is an open source conference geared towards practioners and not a 3-day sales pitch.

There are a few CNCF projects that exist that could be considered serverless but I want to look at two in particular, Knative and CloudEvents. I had the chance to meet some of the great contributors and maintainers of Knative such as Evan Anderson and Calum Murray. It was nice to talk to them about the future state of Knative.

It was equally fun to see Coreweave mention their use of Knative when building their architecture (full talk here). I purposefully am calling this out as Coreweave is one of the most promising startups right now when it comes to Cloud Native AI at Scale. Their use of not just Kubernetes but Knative to serve their inferencing layer is promising and validates the ability to use it.

There was bunch about WASM too so let’s dive into that…

What’s new with WASM?

WebAssembly (aka WASM) is still making big news. I mentioned KubeCon and WASM was a big topic. I had the chance to listen to Jiaxiao Zhou’s KubeCon talk titled “Running WebAssembly (Wasm) Workloads Side-by-Side with Container Workloads”. I am not going to spoil it but he shows some really cool information about ways to run WASM on containerd in a Kubernetes cluster.

In the past I have talked about how one of the shortcomings of WASM was it’s lack of an orchestrator like Kubernetes. Seeing that it can potentially use k8s now, this is promising for the future. Again, I don’t want to give away too much and encourage you to watch the linked video.

Now some other WASM news, Fermyon donated SpinKube to the CNCF. What is SpinKube? Well it is a tool that simplifies building WASM-based microservices applications on top of Kubernetes.

This story dovetails nicely with the earlier one. WASM is incredibly promising but for it to work at scale, we need a way to orchestrate the microservices. There have been several projects dedicated to doing just this but I like that this one has been donated to CNCF.

Historically, the open source community has been amazing at creating tools that revolutionize tech. After all, NGINX and Apache are the most popular web servers while Unix based OSes (largely BSD and Linux) remain popular for server operating systems. Android is the most popular mobile operating system, the list goes on and on.

I think that by sharing this problem with the open source community, we will be able to create a better solution. I do think WASM will exist side-by-side with containers in the near future. It’s still early now but it’s likely to happen. I don’t think it will surpass containers in the same way that containers didn’t eliminate VMs. They will just be two runtimes that people can use.

According to The New Stack, platform engineers are embracing WASM for serverless applications. These packages are smaller and promise to be faster.

Speaking of platform engineering.

Serverless Platforming Engineering

Look, AI is takng over the world. No I don’t mean in a Terminator/SkyNet way. You just can’t have a conversation about tech without talking about AI. It is impossible to build a product strategy nowadays without including AI in that strategy. It’s here to stay.

Another concept that has really taken off in the past 2-3 years is “Platform Engineering”. It seems to be the spiritual successor of DevOps. I honestly heard next to nothing about it in 2022 and then in 2023 and 2024, I hear it everywhere.

In short, platform engineering is buildling toolchains and workflows to help developers be more efficient and allow them to self-service. Basically, it is figuring out ways to simplify the deployment process for developers. How is this different from DevOps?

DevOps, in short, is the practice of the development teams and operator teams working in tandem to deploy an application from code to production. This usually involved creating a CI/CD pipeline to streamline the build and deploy process of the software development lifecycle but also ensuring that everything is secure, workflows are continuously being improved and that both teams are working together.

According to The New Stack, platform engineers earn more than “traditional” devops engineers. I think it’s safe to say that Platform Engineering as a concept has some staying power. But hey, if the goal is to make deployment easier for developers… why not use serverless?

The challenges around platform engineering with serverless is building a fully integrated platform. In the CNCF, there are tools like the aforementioned SpinKube that could be used to simplify WASM development. Tools like Knative ARE tools that enable you to build a platform on Kubernetes to let users self-serve. But how do we connect all these pieces?

Existing tools like GitLab can help address this and Dagger looks very promising. This provides the primitives to simplify the workflows. Backstage is promising but I also think it’s early in it’s lifecycle and there are some feature gaps that I’d like to see addressed. ArgoCD on a cluster could also do some cool things.

As I see more news on platform engineering for serverless, I will be sure to share it here!

Happy Birthday Lambda (and FaaS)

I am not a FaaS fan. That should be no secret. I think it serves a great purpose BUT I don’t like how FaaS has become synonomous with “serverless”. But, I also need to give credit where credit is due. Lambda invented a space that didn’t exist before and really got the serverless compute movement going.

So with that I say….

Happy Tenth Birthday AWS Lambda!

In my blog post on the history of serverless I talk about my friend Mark and I being at SxSW in 2015 and hearing about the launch of Lambda. To see how much it shaped the market is astounding and we have to give credit to one of the OGs of cloud serverless compute.

Closing Thoughts

There were so many talks about serverless this past month with KubeCon/CloudNativeCon and just in general. We have had modern serverless for a decade now and serverless containers are growing in popularity along with WASM!

Now as people adopt platform engineering, I anticipate that we’ll see more and more serverless adoption as that would be the best way to help developers self-serve!

—Photo courtesy kaboompics.com on Pexels—

Day after day I hear people confuse Function-as-a-Service (FaaS) and that just tells me that I have more work to do.

Now you may notice that I have gone down to one post a month. Life has gotten a bit busier so I apologize for that. However, I have some good news. In addition to this newsletter, a podcast will be launching before the end of 2024. More details to come soon.

So moving forward, we will have one monthly newsletter with a few “mini-posts” here an there as well as the podcast. It is our way of finding ways to continue to give you value in multiple formats.

Anyway, let’s talk about some of the top serverless news in October of 2024!

Wiz now supports Serverless Containers for Security Scanning

Wiz is an amazing security company. Container security has always been a challenge and while there are some ways to accomplish this, there haven’t been a lot of third-party companies that specialize in this area. Aqua Security has always been a contender and they do great things but I think it’s important to have competition.

Wiz was founded in 2020 and within 4 years Google offered around $23B for acquisition and Wiz rejected it! That shows not only how fast they grew and the impact they had but also their leaderships’ confidence in their offering. They are pursuing an IPO.

Anyway, it looks like Wiz is now offering security for serverless containers. This is a big deal as it shows that companies like Wiz are seeing serverless containers as a viable option in the marketplace and want to secure them.

Datadog did a Container Report at the top of 2024 and they noted the growth in serverless containers. By September 2023 just under 50% of organizations were leveraging serverless containers.

Cloud security will always be something that is incredibly important. Can you even go a month without hearing about some kind of crazy security breach? I think some people get complacent with security by thinking “well we are on the cloud so clearly we are secure because X provider is securing things”. Security is a shared responsibility model. You are required to secure your applications and services. Tools like Wiz can help.

OpenPR lists a story about the expected surge of serverless security moving foward. By extending to support serverless containers, we will likely see more adoption as more people will be comfortable.

AWS Redshift Serverless is now in AWS GovCloud.

It should be no secret that the government has high expectations of third parties providing cloud services. They take security very seriously and anyone who has worked with the federal government, the RFP process is a pain.

AWS, like every cloud provider, offers a GovCloud which is a special region to better isolate government data. Well now, RedShift Serverless is on AWS GovCloud. Why does this matter? Well as I stated, the government has high standards for cyber security. They will not use anything that doesn’t meet their standards.

AWS, while a competitor to my employer, is a serious enterprise. They are not going to place a product in their GovCloud offering that doesn’t meet the security standard otherwise they risk the future of their business. This tells me that AWS not only made the investment to make RedShift Serverless compliant, but they saw the investment as either necessary (people are asking for it) or worthwhile (people may not be asking for it but will like it when they see it).

Having a serverless data warehouse that the US government sees as trustworthy enough to store data and perform analytics speaks volumes to the growth. Now the offering is new so it is unknown to me at this point how much it will be used, but again, they wouldn’t invest in it if they didn’t at least see potential.

Huawei Cloud talks Cold Starts

Huawei Cloud is not one we hear about often here in the West. In fact, I don’t think they even crack the top 10 cloud providers globally. They are huge in China and have datacenters in something like 33 regions.

They even grew to $7.6B in revenue in 2023. So they are a real contender in the space even though we don’t hear about them often in the USA. Honestly, this is the first time I really heard about Huawei Cloud. I always saw them as a phone and networking device manufacturer.

I was elated to realize that they recently shared some research that they did with serverless runtimes. They actually talk a bit about serverless containers.

The main call out is the “cold start problem”. To recap what this means, when a serverless application scales down, it’s not running at all. So when the URL gets hit by a user or API call, it will take a moment for the service to “wake up” and respond. This is often only milliseconds but can sometimes be a few seconds long.

This is one of the major trade-offs with migrating to serverless. Many people have tried to engineer around it and offer “warm starts”. The service is sort of “awake” but not really so the startup time is much faster.

They don’t provide a ton of alternative options to address it but they do call out the trend. They also go into deeper details about the various contributors to cold-start issues such as runtime language and triggers.

The data and research is on GitHub so I recommend you taking a look.

Vultr gets into the Serverless AI game

Vultr is the world’s largest privately held public cloud. They just announced their serverless inferencing platform. This is a part of the larger “Inference-as-a-Service” that I have mentioned previously.

We are seeing the rise of “Agentic AI”. Generative AI (GenAI) gives you the ability to ask an LLM a question and get a response. However, we aren’t going to ask people to use a CLI or something to do that. They expect a user-friendly interface and chat are the common way to accomplish this.

Well, the code for that bot needs to live somewhere. This is where serverless is PERFECT. The chat bot application scales down to zero when it’s not being used and then kicks on and does what it needs to do to “chat” with the user and access the LLM then scale down.

This also helps with edge computing. Speed is going to be important with chatbots so being able to have that code exist as close to the end user as possible is important. Serverless applications tend to be lightweight which makes them amazing.

Another great thing is that it will come with a TurnKey RAG for customization. After all, the LLM isn’t trained on your data (at least I hope not) but you want it to give relevant information. RAG is the best way to accomplish this.

This report from BusinessWire is a good read on this. What Vultr is doing with Serverless is very impressive.

Closing Thoughts

Serverless is here to stay. That’s just a fact. AI and edge computing have become a sort of tipping point for this.

We are seeing security companies seeing it as a viable option so they want to corner the market for it. We are also seeing the government possibly being interested in using it. It is being used for inferencing with LLMs and other AI models.

We are also seeing companies investing in research and development to address known issues with serverless such as cold-starts. Serverless is growing and is providing a lot of value to developers. But as this Information Week article suggests, we need to think about toolchains and how to best leverage this platform for teams. By having conversations around architecture, we may be able to best address this.

—Photo courtesy Toni Cuenca on Pexels—