IntermediateSITUATIONAL
You inherit a large legacy SPA with poor test coverage and frequent regressions—what immediate steps do you take in the first 30, 60, and 90 days to reduce risk while continuing feature work?
Frontend Developer
General

Sample Answer

When I took over a 5-year-old SPA used by ~200k monthly users, regressions were slowing releases weekly. In the first 30 days I focused on triage: set up monitoring (Sentry + Lighthouse), identified the top 10 flaky routes causing 70% of errors, and added smoke tests around those flows. For 60 days I introduced a testing pyramid: lightweight unit tests for core utilities (covered 40% of hot paths), component snapshot tests for critical UI, and 10 end-to-end tests that ran in CI and caught three regressions before release. By 90 days I prioritized a continuous refactor plan—modularized two large bundles, reduced a critical bundle size by 30%, and established a pull-request checklist and gating so new features shipped with tests. This cut regressions by ~60% and sped release cycles from weekly patches to biweekly planned releases.

Keywords

Prioritize monitoring and triage to focus efforts where impact is highestIntroduce a pragmatic testing pyramid: smoke, unit, component, and targeted E2EShipable governance: CI gates, PR checklists, and incremental refactors