Platform engineering is the process of planning and implementing toolchains that enhance software development and delivery. Platform engineers handle tasks that help application developers work more efficiently, such as preparing CI/CD pipelines, setting up staging environments, and configuring Infrastructure as Code (IaC) to automate cloud resource provisioning.
In this article, you’ll learn what platform engineering is and how it differs from adjacent concepts, including DevOps and SRE. We’ll cover some best practices for implementing platform engineering and explain how they benefit your software workflows.
Platform engineers create new tools and workflows for developers to use. They produce an integrated environment for building, testing, and iterating upon changes. The “platform” bridges between developers and your infrastructure, the scope of which might not be feasible for individual developers to reproduce.
Self-service access is a defining characteristic of effective platform engineering. Developers should be able to consume the capabilities they need for their work without relying on other teams, such as Operations and Infrastructure, each time.
Having to seek approval to spin up a new staging environment is a bottleneck that impedes efficient development. Conversely, running a terminal command that instantly starts an isolated environment empowers developers to be more autonomous. They can stay productive without waiting for infrastructure or having to understand how it’s provisioned.
The Principles of Platform Engineering
Platform engineering combines several technologies and methodologies to produce a holistic development experience:
- Automation and IaC – Infrastructure should be automated and reproducible. IaC tools are used to define what the platform should look like and enable new instances to be created on-demand. Manual actions are kept to a minimum to eliminate friction points in the development flow.
- Focus on efficiency – The platform should be designed to solve the most common challenges encountered by developers. Focus on supporting the unique needs of your teams, instead of trying to recreate functionality that already works well. Rolling your own CI/CD or source control system is unlikely to be beneficial, for example, but providing a mechanism that mirrors a snapshot of your production infrastructure into a fresh staging environment could save developers hours each week.
- Self-service access – Every part of the platform should be an asset that developers can freely utilize. You’re providing a toolbox of controls for developers to use as they see fit. Avoid prescribing specific usage patterns, as individual engineers may work in slightly different ways.
- Continual evolution – The platform should be continually developed using the same product-driven mindset you apply to customer-facing functionality. The “customer” happens to be internal developers, but it’s still vital that improvements are implemented promptly, so the platform effectively meets their needs. This ensures developers stay productive over a sustained period of time.
Each one of these principles revolves around simplifying the development experience. It’s the job of platform engineers to listen to developers, then provide the toolchains they require.
The Platform Engineering Role
Platform engineers have several responsibilities. They’ll discuss challenges with developers and then act upon their insights to build internal platforms. This can include the following tasks:
- Configuring IaC tools to provision new infrastructure on-demand.
- Working with existing infrastructure and operations teams to “narrow the gap” between dev and prod.
- Implementing and maintaining CI/CD pipelines that automate inefficient workflows.
- Creating bespoke internal tools to accommodate org-specific workflows, enforce security policies, and maintain compliance with regulatory standards.
- Building, maintaining and documenting custom APIs, CLIs, and web UIs that expose the platform’s functionality. This could be an API that exposes the number of errors in each environment or a CLI that pushes local code straight to a new sandbox.
The work often culminates in a form of internal Platform-as-a-Service (PaaS). Developers get to deploy their applications onto infrastructure using a fully automated workflow that requires no specialist knowledge. The platform coordinates complex procedures such as creating cloud resources, deploying containerized microservices, setting up networking, and seeding any test data the developer requires.
Organizations can be hesitant to invest in platform engineering. A common concern is whether the engineers working on the platform would be better utilized within the main product team. Here are three reasons why you should commit to platform engineering.
Internal platforms accelerate development. Automated processes and self-service infrastructure help to keep developers productive. They’re able to continually move forwards on the product features prioritized by the business.
Once a feature is ready, the development team can spin up a new test environment to autonomously verify the change. The platform can perform automated tests and then ship the feature to customers in production while developers start work on the next task. This reduces your time to market without compromising on quality.
Promotes Focus and Specialization
Developers should be able to focus on what they’re best at: development. Modern infrastructure, CI/CD, and distributed deployment systems are dedicated specialisms. Developers don’t need to be experts in these fields and may sometimes struggle to understand them.
Adding platform engineers lets developers stay productive by concentrating on building new software. The platform engineering team can be staffed with experts skilled in relevant topics such as IaC, CI/CD, and PaaS solutions. Advancements in both development and infrastructure will proceed more quickly as each individual will be specialized in their role.
Ensures Tools and Processes Continually Develop
Development processes need to evolve as your product grows. Over time, your stack expands with additional technologies and new requirements. You might introduce a new storage system, require more comprehensive end-to-end tests, or have to comply with another regulatory standard.
Platform engineering ensures your toolchain develops in tandem. Without it, you have to make ad-hoc workflow adjustments which can be poorly documented and difficult to maintain. Moreover, developers often lack the time to make the optimizations they require. This perpetuates the use of inefficient practices after they’ve been recognized as a bottleneck. Providing devs with access to a platform engineering team allows frustrations to be addressed without delaying the product’s release schedule.
Automation, testing, monitoring, and IaC: these characteristics are already familiar to DevOps practitioners, so what sets platform engineering apart?
First off, platform engineering shouldn’t be viewed as an alternative to DevOps. It’s more accurate to treat it as an implementation of DevOps concepts and philosophies. The overarching aim of DevOps is to simultaneously improve software quality and throughput using new tools, processes, and collaboration frameworks. Platform engineering is an example of what this looks like in practice.
Providing developers with self-service access to infrastructure shortens the feedback loop and reduces complexity. This enables more focused work on forwards-facing tasks relevant to your business aims. Platform engineering accelerates the development cycle, which achieves the objectives expressed by DevOps.
In summary, platform engineering sits separately to DevOps but is usually part of a DevOps strategy. Seen from the other side, DevOps is more than just platform engineering, as a complete DevOps flow will extend beyond internal development tasks to deliver and manage code in production.
Platform engineering neighbors Site Reliability Engineering (SRE) too. The main purpose of SRE is to preserve the stability of your production environments. These teams use objective data-driven targets such as SLAs and SLOs to identify when incidents materially affect your customers or your business. SRE then manages the incident resolution, analyzes what went wrong, and implements changes to prevent the problem from recurring.
Because platform engineering looks at internal systems, it doesn’t directly overlap with SRE. SRE produces infrastructure that’s optimized for highly reliable operations. Platform engineering creates assets that facilitate high-velocity development.
Information should be shared between the disciplines, though, as insights from one field are often valuable to the other. Difficulties encountered while setting up an internal workflow could reveal opportunities to simplify production infrastructure, for example. The idea is not to silo off the concerns but instead keep them as complementary philosophies that you can iterate upon.
Platform engineering is the practice of building and maintaining internal toolchains that support software delivery workflows. A dedicated team implements tools and processes that provide developers with self-service access to infrastructure.
Companies that pursue platform engineering can acquire a competitive edge in their field. Developers are able to work more autonomously without constantly waiting for approvals from infrastructure admins. This means more time spent adding new features to your product.
Platform engineering doesn’t mean you have to create everything from scratch. Take a look at Spacelift to start building the internal platform you need. Spacelift offers collaborative CI/CD for your infrastructure and workflow automation, letting you unlock developer freedom while retaining precise security guardrails.
The Most Flexible CI/CD Automation Tool
Spacelift is an alternative to using homegrown solutions on top of a generic CI. It helps overcome common state management issues and adds several must-have capabilities for infrastructure management.