Continuous Localization in Enterprise Web Projects
summary
Company migrates localization to an automated pipeline that scans 119 git repos across 19 micro-frontends and 40 teams, pushes strings to a cloud TMS, pulls translations, publishes language bundles to CDN with cache invalidation, and generates PRs for English fixes. Workflow relies on using English source phrases so missing translations fall back automatically, translation memory for reuse, and daily scan and build jobs. A lightweight QA feature called fake-language coverage exposes untranslated UI and surfaces UX-writer edits.
AI-written summary · about
chapters (19)
- 0:00Introduction by host
- 2:57Speaker introduction
- 4:25Why localization matters
- 10:13Key localization concepts
- 13:04Enterprise use case at Infobip
- 16:35Problems with manual translation
- 21:11Automating translation workflows
- 23:00Tools for translation management
- 24:47Building automation at Infobip
- 30:03Integrating localization into CI/CD
- 31:26Workflow deep dive
- 33:57Scan and build jobs explained
- 38:17Rendering localized UI
- 40:25Fixing translation typos
- 41:53UX writer tools
- 42:15Tracking translation coverage
- 43:20Collecting translation feedback
- 44:22Lessons learned
- 46:28Q&A and closing
related talks
26mContinuous Localization in Enterprise Web Projects
21mContinuous Translation: I18n in Agile Projects
23mAutomating Localization
23mDriving Seamless Localization
12mHow we automated code maintenance
04mGlobalization Pipeline: Translations at cloud speeds
20mFinding translations, localisation and internationalisation
34mMass Frontend Refactors: Less Stress; More Success