Using ML to find value in your automated tests
summary
Open-source dashboard ReportPortal.io collects test results in real time, tags failures, and uses TF‑IDF + k‑NN lazy learning over vectorized stack traces (stored in Elasticsearch, analyzed by a small Go service) to auto-categorize fails as product, automation or system issues. It shows live dashboards, links to JIRA/Sauce Labs, and integrates via API/Jenkins. Community adoption hits ~300k Docker pulls; ML starts helping after ~10–15 runs (much better after hundreds). Current limits: 150 launches on MongoDB; v5 moves to PostgreSQL and raises that.
AI-written summary · about
chapters (10)
- 0:00Start
- 3:58What Is the Competitive Center
- 6:54Report Portal
- 9:01Real-Time Reporting
- 9:34Open Api
- 9:59Integrations
- 12:16How It Works
- 23:38Normalization
- 37:45Elasticsearch Setup
- 39:14Is There a Way To Push Test Results in a Predefined Test Execution Which Is Created by Api
related talks
40mMachine Learning helps to work with your test automation
41mReportPortal.io – How to Make AI Analyze Your Test Automation Reports?
45mBuilding loose-coupled continuous testing platform
32mLightning Talks
51mReportPortal.io - visibility of test automation status, metrics...
39mMethodology of ML in Performance Testing (How to Find Bottlenecks Automatically)
31mUse-Case Driven ML Platforms
46mSW Testing: Can ML save us?