You are the Engineering Manager for a Databricks Lakeflow team responsible for orchestration, pipeline observability, and developer productivity. Your team supports a fast-growing enterprise customer base, but the next 2 quarters contain more committed roadmap work than the team can realistically deliver. Leadership wants a clear recommendation on what to build first, what to defer, and what to cut.
The team has 8 engineers, 1 EM, 1 PM, and shared support from Design and Product Marketing. You have 16 weeks before the Databricks Data + AI Summit, where the VP wants to announce meaningful Lakeflow improvements tied to customer adoption. The problem: five major roadmap items are competing for the same capacity, and several have executive or customer pressure behind them.
The VP of Engineering wants a launchable story for Summit. The PM wants to improve activation and retention for mid-market users. The Field CTO is pushing for two enterprise commitments tied to customer renewals. The Staff Engineer is concerned that shipping new features without platform hardening will increase incident load.
You have capacity for roughly 32 engineer-weeks of roadmap work after accounting for on-call, maintenance, and incident response. The candidate roadmap items are estimated as: 1) Lakeflow pipeline templates (8 engineer-weeks), 2) cross-workspace run visibility in the Databricks UI (10), 3) SLA alerting for failed jobs (6), 4) Unity Catalog lineage integration for pipelines (12), and 5) migration tooling from legacy Jobs orchestration to Lakeflow (14). Budget allows at most 1 contractor for QA support. Two engineers are allocated 25% to production support, and one senior engineer is unavailable in weeks 7-9.