Serverless compute: code that runs without a server you look after
You hand over a function and the conditions under which it should run. The platform starts it, scales it to whatever the traffic needs, and bills you for the time it actually ran.
What is serverless compute?
There is still a server. You just never name it, patch it or pay for it while it sits idle. You write a function, say what should trigger it, and the platform handles everything between the trigger and the code: capacity, scaling, restarts after a crash.
The trade is in the shape of the work. Short, bursty, event-driven jobs fit it perfectly and cost almost nothing when nobody is using them. Long-running jobs, work that needs a GPU for an hour, or anything that must hold state between requests fits it badly, and the bill says so.
Four ways to run it
Each cloud sells the same idea under its own name, and there is an open implementation you can run on any Kubernetes cluster.
AWS Lambda[1]
The original of this category. It runs a function per event, scales out to many copies at once and charges per millisecond. It is the piece most of my AWS architectures are glued together with.
Official documentation →Azure Functions[2]
The same model on Azure, with a choice of hosting plans that decide whether you pay only for what runs or keep instances warm for predictable start-up times.
Official documentation →Cloud Run functions[3]
Google folded its functions product into Cloud Run, so a function and a container run on the same platform and you can move from one to the other without changing clouds.
Official documentation →Knative[4]
Scale to zero and event-driven triggers on your own Kubernetes cluster, including an OpenShift one. It is the answer when the code may not leave your building but you still want the serverless shape.
Project documentation →Where I used it
The video pipeline
Every render job and every small piece of glue between the steps is a Lambda function. A batch of clips finishes in about the time one of them takes, because the same function simply runs many times at once.
Read the write-up →Agent tools
When an agent is given a tool, the tool is usually a function behind an endpoint. Serverless is the cheapest way to hold a tool that may be called a hundred times in a minute and then not at all for a week.
AI agent runtimes →Managed and self-hosted together: hybrid
Serverless is where hybrid gets decided in practice, because the function is the thing that touches the data. The usual split: the orchestration and the public-facing functions run in the cloud, and the one step that reads the actual files runs on Knative inside the building, triggered by the same event.
What would help you next?
Learn to decide
The course “From Managed to Hybrid” teaches where AI should run: what each layer costs, what you operate and where your data has to stay. It is in preparation, with a free live session first.
Linda MohamedAI and cloud consultant, AWS Community Hero, Vienna
Build it with me
Tell me what you want to build. In 30 minutes we check which of these ways fits your data, your rules and your budget, and whether AWS funding for a proof of concept can cover part of it. How AWS funding works →
You already know what you need
Book the workshop directly:
- AI Concept and Prototype workshops →
Architecture first, priced per service and tested on your data, then a prototype with a go or stop decision. - AI Discovery Workshop →
Sort your ideas into a scored shortlist, so you know which use case to build first.
Read more on ai-solutions.wiki
My open knowledge base, with longer comparisons, pricing notes and sources.
Sources
- AWS Lambda documentation. https://docs.aws.amazon.com/lambda/latest/dg/welcome.html
- Azure Functions documentation. https://learn.microsoft.com/en-us/azure/azure-functions/functions-overview
- Google Cloud Run functions documentation. https://cloud.google.com/functions/docs/concepts/overview
- Knative documentation. https://knative.dev/docs/
- AWS Lambda, ai-solutions.wiki. https://ai-solutions.wiki/tools/aws-lambda/
- Azure Functions, ai-solutions.wiki. https://ai-solutions.wiki/tools/azure-functions/
- Knative, ai-solutions.wiki. https://ai-solutions.wiki/tools/knative/
- Serverless, ai-solutions.wiki. https://ai-solutions.wiki/glossary/serverless/
- What is serverless, ai-solutions.wiki. https://ai-solutions.wiki/basics/what-is-serverless/
Links checked October 2026. Product names and features change: check the linked pages before you decide.
