Technical Program Manager – Program Management

Technical debt programs fail in one of two directions — an unprioritized wishlist nobody funds, or a competing initiative that reliably loses to the feature roadmap. The fix is to make debt visible and quantified centrally, then fund its paydown as a fixed, budgeted slice of normal planning instead of a side project asking for favors.

1. 1.Establish a shared definition and taxonomy — split debt into categories (security/reliability, scalability, maintainability, deprecated dependencies) with different urgency profiles, and scope the program to shared, foundational cloud components where debt has real blast radius.
2. 2.Build a centralized, scored inventory — pull debt items from architecture reviews, postmortems, dependency scanning, and static analysis, and score each on blast radius, cost of delay, and remediation cost.
3. 3.Assign ownership through a lightweight governance structure — give every foundational component a named owning team accountable for its debt score trending down, with a small cross-team architecture council arbitrating ownerless, cross-cutting items.
4. 4.Fund paydown as a fixed capacity allocation, not a competing project — have every team allocate a standing 15–20% of sprint capacity to debt paydown, with periodic hardening cycles reserved for large, high-scoring items that can't be chipped away at incrementally.
5. 5.Prioritize with a cost-of-delay framework, not volume or recency — rank the inventory by a WSJF-style score and back the top items with real incident and velocity data so prioritization is defensible against a feature asking for the same capacity.
6. 6.Instrument debt continuously and run a standing review cadence — wire static analysis, dependency scanning, and SLO error-budget burn into CI so new debt is scored as it's introduced, and report debt-score trend by component to leadership every quarter.

This holds up because debt paydown is funded the same way feature work is — a fixed, budgeted allocation inside normal sprint planning — rather than a separate initiative that has to win a roadmap argument every quarter; the scoring model and quarterly trend review keep that budget pointed at the highest-leverage items instead of whatever's loudest.