DragonCrawl: Revolutionizing Mobile Testing with GenAI
summary
An LLM-based end-to-end testing system that simulates mobile UI semantically by extracting screen state, sending actions as natural-language tasks, and continuing until a success screen; originally deployed at Uber and caught $25M in four months by detecting 11 P0 production bugs. It runs on web, Android, and iOS, converts session replays into tests, ties UI runs to backend traces to pinpoint the failing microservice, and aims to cut the 30–40% engineer time spent on brittle selector-based test maintenance.
AI-written summary · about
chapters (7)
- 0:00Introduction and background
- 1:46Problem with manual testing
- 3:11DragonCrawl solution at Uber
- 4:59Tracing and debugging benefits
- 6:26AI codegen and testing demand
- 7:29High test maintenance challenges
- 8:13Q&A on testing integration
related talks
29mDragonCrawl: The AI Testing Framework That Adapts Instead of Breaks
35mApp Testing with AI Tools – Make it Easy
19mQA Agents for Android: How AI Is Changing the Game in Testing
18mReinventing App Testing with Agentic Apps
19mThe GPT-ification of Mobile App Testing
49mPanel Discussion: AI in Mobile QA - What Actually Works Today?
23mReimagining Mobile App Testing with Vision AI-Powered Agents
41mAI Driven Mobile Testing