SiriusXM's platform engineering team built a data-driven prioritization framework that treats assumptions like code — documenting, versioning, and reviewing them in source control. The framework defines platform-specific impact factors (development speed, reliability, security, cost, etc.) and uses a scoring model combining impact, urgency, effort, and business needs. Developer survey data and quarterly reviews keep weights current. AI is layered on top to surface relevant assumptions, connect initiatives to OKRs and prior project data, and generate initial scores — while humans retain final judgment. Each prioritization cycle enriches the knowledge base, making the system more valuable over time.
150 Impressions