Scaling Frontend: Supercharging Productivity with Dedicated Infrastructure Teams
summary
A dedicated frontend infrastructure team enables focused work on Ember tooling, CI, and DX for a 9.5‑year monolith of ~700,000 lines with ~150 engineers across 30 teams. Projects include adopting Tailwind, contributing to Embroider, rolling out localization to ~80% of main traffic paths, and balancing Ember exam test buckets via lint rules and max‑tests‑per‑file. CI and test fixes cut p50 end‑to‑end times from 10m to 5m, reduced browser memory leaks, and support ~220 production deploys per week.
AI-written summary · about
chapters (10)
- 0:00Introduction and context
- 3:32Growth challenges in legacy codebase
- 8:02Impact of shipping delays on productivity
- 10:28Creating dedicated frontend infrastructure team
- 14:21Roadmap inputs and prioritization
- 17:07Measuring success beyond metrics
- 23:28Case study: Internationalization
- 25:33Case study: Frontend observability
- 28:21Case study: CI performance improvements
- 31:50Summary and team impact
related talks
31mTaming the Beast: Managing a really ambitious codebase
36mPanel: Ember at Scale
28mModernizing a Massive Monolith
15mEstablishing Platform Teams to Utilize the Economy of Scale
41mMicro-frontends: the golden circle
35mMicro frontend Has Entered the Game
59mEmber 2.0: Out-of-the-Box Productivity
07mLightning Talk