
Databricks is preparing a quarter-end release for Unity Catalog focused on two customer-facing capabilities: faster cross-workspace data sharing and improved lineage visibility in Catalog Explorer. Your engineering team owns the metadata service and permissioning layer behind both features, but the same codepath is responsible for several reliability issues and a growing backlog of technical debt.
You are the Engineering Manager for a team of 10 engineers, 1 TPM, and 1 QA engineer. The VP of Product wants both features available to design partners in 12 weeks because they are tied to expansion deals worth $6M in ARR. At the same time, the on-call burden has doubled over the last two months: the service missed its 99.9% availability target in 3 of the last 8 weeks, and engineers have identified 14 high-priority debt items, including flaky integration tests, slow permission-check queries, and an outdated deployment pipeline.
The Group PM wants to maximize feature scope for the quarter. The Director of Engineering wants measurable reliability improvement and lower pager volume before approving broader rollout. Sales is pushing for committed dates to support renewals, while Support wants fewer incidents and faster mitigation steps.