Test Automation in the Age of Al
summary
Panel reports concrete results and limits: a low‑code approach implements self‑healing via multiple element interactions for flaky UIs, an AI-driven failure grouping and fix recommendations system reduces triage time, and GitHub Copilot generating unit-test stubs speeds authoring. Open tools exist—ReportPortal.io open-source since 2018—but many enterprise products are paid. Claims include QA of over 5,000 ML models and a support case cutting response from 1m4s to instant and resolution from 2h13m to ~3m12s, while governance, token limits and fine‑tuning remain blockers.
AI-written summary · about
chapters (16)
- 0:00AI in test automation
- 1:09Panelists' AI backgrounds
- 4:41AI tools and self-healing tests
- 6:30AI test failure analysis
- 7:32AI optimizing test selection
- 9:39AI test data generation
- 13:38Training data and prompt limits
- 16:52Open-source AI tools limits
- 19:50AI governance challenges
- 21:39GitHub Copilot for tests
- 23:26AI test data and learning
- 26:58AI for documentation
- 30:35Future AI risks
- 33:50AI impact on testing jobs
- 37:00Adapting to AI in testing
- 41:14Closing remarks
related talks
01h 04mImpact of AI on #SoftwareTesting: Are Testers Ready? 🔮
26mMaximizing Test Coverage with AI-Powered Test Automation
06mTriple Threat: AI Technologies Reduce Testing Costs
40mFrom Bugs to Brilliance: Leveraging AI in Modern Software Testing
01h 05mPanel Discussion - The Future of Testing: Impact of AI in Quality Assurance and Beyond
25mWhat’s Our Job When The Machines Do Testing?
25mSelf-Healing in Test Automation
41mHow AI is transforming software testing