In this episode of Tech Chat with Navika Chadha, a Cloud Engineer and Microsoft MVP, we had a deep conversation about Platform Engineering: what it is, how it differs from DevOps, why companies should adopt it, and what skills are needed. The discussion cuts through the buzz and click-bait headlines to explain why DevOps is very much alive and how Platform Engineering complements it.
What Is Platform Engineering?#
Platform Engineering is about enablement. A Platform Engineering team provides a platform as an internal product to the other teams that create products. This platform should contain everything the product teams need to develop their products faster and more efficiently.
The platform team generates value for the product teams, and the product teams generate value for the client. The reason many companies are moving in this direction is simple: the cognitive load on product teams was rising. They needed to handle too many tools, too much technology, too much complexity. There were redundancies across teams. Platform Engineering reduces that cognitive load through standardization.
Platform Engineering enables other teams to do DevOps. It is an enabler, not a replacement.
Why Should Companies Adopt Platform Engineering?#
The benefits flow in two directions. For internal teams (software developers), the platform increases efficiency through automation and standardized tooling. For clients and users, it results in more consistent, higher quality output. It is a win-win situation.
Platform Engineering is essentially about standardization. You give teams a “golden path,” a recommended way of building and deploying software. For about 80% of use cases, the platform should be the right solution. But you always need to allow teams the freedom to break out and do their own thing when necessary.
What Capabilities Should a Platform Have?#
Every platform will look different depending on the company’s needs. But generically, several capabilities appear consistently:
- Application Runtime: Where applications run. This could be on-premise, cloud, Docker, virtual machines, or a Kubernetes cluster.
- Automation: Automating repetitive tasks like certificate rotation, password changes, and vulnerability management (container scanning, SCA, SBOM).
- Observability: Default dashboards, tracing, and monitoring already in place so teams do not need to build that from scratch every time.
- CI/CD Pipelines: Pre-built pipelines with sensible defaults that teams can use immediately for building their products.
The key insight: platforms are mostly about reusable, repeatable tasks. Instead of starting from scratch manually every time, you standardize so teams can focus on what matters most, delivering features.
DevOps vs. Platform Engineering: Complement, Not Compete#
There is a lot of buzz claiming Platform Engineering will replace DevOps. But much of that is click-bait. Here is the reality:
The DevOps movement started by combining development and operations, then evolved toward continuously delivering value by incorporating all people working across the value stream. This moved us from projects to products. But this comes with a price: you need to remove silos, which means the cognitive load on teams rises because they need to handle everything.
Platform Engineering was born as a response to that cognitive load. A platform engineering team uses DevOps practices to build the platform. Product teams use DevOps practices on top of that platform to build their products. DevOps is not dead. It is still the foundation.
Platform Engineering and DevOps complement each other. Platform Engineering is the strong foundation that makes DevOps more efficient. But if you are a startup with one team, you do not need Platform Engineering. You just do DevOps. The Platform Engineering case comes into play when you have multiple teams with redundancies and you want to get faster while bringing down costs.
Platform Engineering Tools#
The tooling landscape is growing rapidly. Microsoft has released tools in this space. OpenShift can be looked at as a platform engineering tool. Backstage is a very good example of a developer portal. Many more tools are emerging, and I expect a Gartner quadrant for platform engineering tools soon.
What is important to recognize: as a company or platform engineering team, you need to look at what your product teams actually need. Is it a standard platform off the shelf, or a custom-built platform assembled on top of other platforms? Usually, companies end up building platforms on top of platforms, stacking capabilities. The key is to identify which capabilities you need, which you can borrow from existing platforms, and which you need to build yourself.
Security and Cost Considerations#
When building a custom platform, all security measures apply. It is an internal product, so you need penetration testing, security testing, license scanning, SBOM analysis, everything. A well-built custom platform can be absolutely secure.
The bigger question is cost. Building a platform requires at least a team of five to seven highly skilled people with 10+ years of experience in software development. They need deep knowledge in Kubernetes, pipelines, security, networking, backups, and IT operations. That investment is significant. But for a company with around 1,000 developers and specialized development needs, building a custom platform absolutely makes business sense. For smaller organizations, using an off-the-shelf platform may be more cost-effective.
Skills for Platform Engineers#
Platform Engineering is a young discipline, but the skill requirements are becoming clear. A typical platform team needs diverse expertise:
- Frontend/UX: Building portal interfaces, often on top of tools like Backstage
- Kubernetes: Deep knowledge of container orchestration is essential
- Security: Understanding vulnerability management, network policies, and compliance
- Pipelines: Expertise in CI/CD automation and integration
- Networking: Knowledge of service meshes, ingress, and network topology
The default profile of a platform engineer is someone with broad Kubernetes knowledge, understanding of software development practices, knowledge of the customer (software engineering teams), and skills in pipeline automation. Usually, a diverse team covers different specializations rather than one person covering everything.
Key Takeaways#
- Platform Engineering is about enabling product teams to do DevOps more efficiently by reducing cognitive load
- DevOps is not dead. Platform Engineering and DevOps complement each other. Platform Engineering provides the foundation, DevOps is the practice.
- Platforms should provide a golden path for about 80% of use cases while allowing teams freedom for the remaining 20%
- The build vs. buy decision depends on team size, specialization needs, and cost. Companies with 1,000+ developers can justify custom platforms.
- Platform engineers need a diverse skill set spanning Kubernetes, security, pipelines, and developer experience
- Every platform should reduce cognitive load and accelerate the time to productivity for new team members