Autonomous Testing. Are We There Yet?
summary
A prototype whoppy autonomous testing platform exists that wires together an assistant an autonomous bot and a brain, stores an app map in Neo4j, collects screenshots and baselines them into a visual regression tracker, and generates random 5‑node journeys to spot visual regressions. It finds real UI issues in simple apps, took over a year for a four‑person team to get working, and reduces human work only after heavy engineering effort so it does not yet eliminate technical complexity.
AI-written summary · about
chapters (17)
- 0:00Introduction
- 2:15Welcome
- 4:10Whats wrong with testing
- 7:03Manual vs automation
- 11:30Benefits
- 14:45How Autonomous Testing Works
- 15:43Features of Autonomous Testing
- 19:45Autonomous Testing App
- 20:47Autonomous Testing Platform
- 21:38Learning and Testing
- 25:12Challenges
- 30:14Summary
- 35:40How to collect enough data
- 37:31Where to find QA
- 38:48Future concept of testing
- 42:22Knowledge
- 44:43Conclusion
related talks
42mThe Path to Autonomous Testing
41mNot Only Cars: “AI, Please Test My App”
47mThe Path to Autonomous Testing
54mAI for Software Testing: The Ultimate Journey
40mDigitizing Testers
31mKeynote: How Software Testing Can Increase Agent Autonomy
18mReinventing App Testing with Agentic Apps
23mReimagining Mobile App Testing with Vision AI-Powered Agents