Test Intelligence for Architects
summary
Introduces concrete techniques and results: use Pareto optimization to reorder tests so a few percent of runtime finds most bugs (6% time → 80% of bugs, 11% → 90%), add test impact analysis using version-history coverage to get even faster feedback (1% time → 80%, 2% → 90%), and run test gap analysis to expose untested changed code. Shows tool TeamScale, a medical-device case where optimized ordering reaches >99% coverage at 8% runtime, and a legacy C system where optimized one-hour gate finds 16 of 31 historical bugs versus 4 by default.
AI-written summary · about
chapters (4)
- 0:00Introduction
- 1:15Session "Test Intelligence for Architects" by Elmar Jürgens
- 44:15Elmar answers questions from the audience
- 50:10Outro
related talks
44mTest Intelligence: Finding More Bugs in Less Time with AI
28mWhen Can You Skip a Test? Tracking Test Impact
37mUnderstanding the successes and the pain points of different testing strategies
51mTesting Software Architecture
42mBeyond “Just Try Harder”- Three effective strategies to tackle bugs
10mBeyond “try harder”: Effective strategies to tackle bugs
01h 15mHow to measure your architecture
21mTEST FOR THE FUTURE