PaaS
PaaS stands for "Platform as a Service": a provider delivers a ready-made environment over the internet in which programs run. The customer only uploads their own program code, while the provider takes care of servers, operating system, and updates.
Anyone who wants to offer a program on the internet needs a computer that runs constantly. In the past, you bought your own machine, placed it in a room, and maintained it yourself. PaaS is short for “Platform as a Service.” A company rents, via the internet, a fully set-up environment in which its own programs can start immediately. The tenant only supplies their program code, meaning the instructions that make up their program. Everything below that — computers, operating system, databases, security updates — is operated by the provider.
Why companies save themselves the trouble of server maintenance
The biggest advantage is saved personnel. A dedicated server has to be monitored, updated, and, in an emergency, repaired at night. That requires specialists, who are expensive and hard to find. With PaaS, the provider does this work for thousands of customers at once. A small team can thus run a service that previously required an entire department.
The second advantage is speed. A new project can go online within minutes instead of waiting weeks for ordered hardware. If the number of users suddenly grows, the platform automatically provides more computing power. If it drops again, you pay less again too. This flexibility is especially important for young companies that cannot predict their needs.
The downside is dependency. Anyone who tailors their application heavily to the quirks of one provider will later find it hard to leave. Experts call this lock-in. In addition, the provider decides which programming languages and versions are allowed. And if its data centers go down, your own service goes offline too.
The layers beneath your own code
You can imagine cloud offerings as a stack of layers. At the bottom lies the hardware, above it the operating system, then the tools for running programs, and at the very top the finished application. With PaaS, the provider draws the boundary right in the middle of this stack. It is responsible for everything up to the tools, while the customer is responsible only for the topmost layer.
In practice, it works like this: the developer sends their code to the platform, usually with a single command. The platform checks which language and which additional components the program needs and packages everything into a standardized runtime environment. It then starts the program on multiple machines simultaneously and distributes users' requests across them. If a machine fails, the program is restarted elsewhere without anyone having to intervene.
It helps to distinguish PaaS from its neighboring terms. With IaaS, meaning “Infrastructure as a Service,” you rent only the bare virtual machine and set everything up yourself — more control, more work. With SaaS, “Software as a Service,” you rent a finished program like an online mailbox and write no code at all. PaaS lies exactly in between. A common analogy: IaaS is the empty kitchen, PaaS is the kitchen with a stove and ingredients, SaaS is the food you ordered.
PaaS in products and stock market news
Well-known offerings include Google App Engine, Microsoft Azure App Service, Heroku, Vercel, and Red Hat OpenShift. When a website or a mobile app responds quickly even though the company behind it is small, such a platform is often behind it. Many online shops and booking systems run this way too.
In the world of AI, the term has become relevant again. Providers sell platforms on which you can train or run language models without buying graphics cards. Such services are sometimes called “AI PaaS” or MLOps platforms. The principle remains the same: the customer brings the data and the model, while the provider owns the heavy technology.
In financial news, PaaS revenues show up in the quarterly figures of Amazon, Microsoft, and Google. These cloud divisions often grow faster than the rest of the business and are considered especially profitable. A common misconception is that PaaS is automatically cheaper. For very large, steady workloads, owning your own hardware can be more cost-effective — which is why some companies are moving back out of the cloud.
