Do you really need platform teams?

Platform engineering is one of the emerging trends in the tech. It promises to increase developer productivity and reduce cognitive load. Platform engineering focuses on abstracting the complexity of managing infrastructure, CI/CD, and observability through Internal Developer Platforms (IDP).

However, having worked in both a startup and a large bank, we experience platform engineering as a burden. The promise of empowering engineers sounds good, but the fallacy often lies in the execution.

The problem we have seen is that platform teams are too detached from the customer. This leads to a lack of understanding of the problems engineers face. The result is often a one-size-fits-all solution. Of course, this can never work, and it forces engineers to create hacky solutions to solve their problems within that IDP. It further limits their freedom and creativity and leads to highly centralized knowledge.

In addition, reliance on platform teams causes delays in development and innovation because they cannot keep up with the speed of open source and AWS offerings. This can even lead to security issues due to late dependency updates. Platform teams are often overworked and have a large backlog, which makes the wait even longer.

We have seen how easily a startup can achieve the same level of maturity by leveraging the cloud. With AWS, there is very little need for platform engineering because AWS provides a platform out of the box.

Platform teams are not unnecessary, but we do question the trend around them. We will show where platform engineering adds most value and how AWS, open source, and a strong community minimize the role of platform teams.