Your E2E Tests Are Lying to You
summary
Gobi finds that conventional E2E suites generate three real bug finds versus roughly 56 maintenance failures, so the team replaces brittle Maestro/Appium flows and text selectors with an AI agent, Autasana. The agent tolerates copy and layout changes, auto-scrolls to offscreen items, performs exploratory clicks to reproduce a missing event bug, flags inconsistent copy between screens, and surfaces logs and performance spikes, reducing test upkeep while catching usability and real functional regressions.
AI-written summary · about
chapters (9)
- 0:00Introduction and setup
- 2:22Why end-to-end tests fail
- 4:12Four eras of mobile testing
- 6:01AI agent testing demo
- 11:15Humans vs AI testers
- 13:44AI catching UX inconsistencies
- 14:26AI usability testing demo
- 15:51Performance monitoring in tests
- 16:12Team acknowledgments
related talks
24mAutomated Mobile E2E Testing
45mMaintaining E2E Test Automation as We Transition from View…
18mReinventing App Testing with Agentic Apps
19mMaestro: Reliable E2E testing made easy
43mAutomated Tests Are Not Enough
35mApp Testing with AI Tools – Make it Easy
34mBreaking the Build Before It Breaks You: The Magic of E2E Tests
21mYour Tests Lack Vision: Adding Eyes to Your Automation Framework