Optimizing Babbel’s Cloud Infrastructure through mkdev’s Staff Augmentation

Project

Babbel is one of the leading language learning platforms with millions of subscriptions sold worldwide. Serving millions of learners 24/7 around the globe requires a fundamentally solid cloud architecture and a complex engineering foundation, which Babbel's experienced infrastructure team handles excellently by leveraging AWS and serverless technologies. Just as an example, Babbel's cloud infrastructure successfully handles multifold traffic increases on a daily basis, usually caused by different events such as marketing promotions or even meeting the increased traffic needs during 2020. Babbel reached out to mkdev for support, with the goal of expanding the development capacity and making further infrastructure enhancements.

Challenge

With numerous ongoing projects, including releasing new products and product features, Babbel's infrastructure team frequently needs to prioritize supporting key company initiatives over working on technical improvements. Meanwhile, technical debt can be filling up the backlog and some of the desired improvements and refactorings have to wait in a queue for a while or alternatively they can be worked on at the risk of slowing down the business-critical implementation work.

Solution

In order to fully concentrate internal resources on the key business projects, Babbel reached out to mkdev to help them tackle some of the larger technical debt projects. This allowed Babbel to reduce the backlog size while keeping the existing infrastructure team lean, focused and avoiding overstaffing.

Working with internal staff, our experts outlined new architecture for some of the cloud components and environments, and then completely took over the execution, including coordination and alignment with other engineering teams to perform critical production deployments.

For example, the mkdev team took over the preparation of internal tooling and deployment processes for the future AWS multi-account setup. This provides the company with the required foundation for the future migration as well as for introduction of new applications and environments.

Another example is a major refactoring of Babbel's service-to-service communication, which resulted in offloading 30% of all traffic from a highly monolithic API Gateway to a new loosely coupled setup, which among other things and in a true platform engineering fashion, made service ownership clearer and provided application teams with new capabilities.

In both cases, the work was done with full coordination with Babbel's infrastructure team, while still allowing the engineers to work on the company’s key projects.

Conclusion

In many mature software companies like Babbel, internal engineers handle all cloud infrastructure space projects. However, unplanned spikes in work volume can present a resource challenge.

The first option to solve this problem, expanding the internal team with more permanent colleagues, runs the risk of overstaffing, while the second option, temporarily expanding the team with additional external experts and focusing their efforts on the projects that would otherwise be difficult to prioritise and complete, provides a true cloud and elastic way forward.

Babbel's infrastructure team chose the second option, with many benefits gained on multiple assignments and making an already solid cloud platform even more flexible and ready for further expansion.

Do you also realize that your infrastructure could use an improvement and your team could use more free engineering space? Give us a call and we'll discuss it!