
SiriusXM's approach to platform engineering prioritization is the focus, highlighting their data-driven framework that evolved from the RICE model. Eleanor Millman and Mina Tawadrous explain how they incorporated urgency and various impact factors like development speed, runtime reliability, and cost reduction, weighting them based on leadership signals. A key element is "assumptions as code," a system where developers use coding agents to validate assumptions against a central repository, resolving conflicts through a PR process. This framework facilitates cross-team alignment and enables data-informed decisions, such as reducing runtime reliability focus when quality scores were high, or prioritizing code review projects based on code review scores. They also discuss AI's role in scaling these processes and enabling citizen developers.
Sign in to continue reading, translating and more.
Continue