Risky business: taking risks in production

Being an engineer, we all have at least one common thread: We like to build things. That is why writing code and architecting platforms that scale for millions of customers is appealing. However, shipping to production and delivering impactful features for customers can be a complex problem to solve; especially when the platform itself consists of tens or hundreds of components or services all used to deliver the final product. With this, we sometimes have to take risks. Taking risks is a common attribute for all teams and platforms in production. How do we know if we’re taking enough risks? How do we know if we’re not taking enough? When taking risks, what are some techniques for moving faster but keeping risk manageable (i.e. A/B Testing, metrics, canaries, alerting, etc). We’ll dive into these questions and examples of where taking risks paid off (and didn’t pay off..) both within our organization and personal careers.