Automation of the Automation using GenAI
summary
Generative AI calls the ChatGPT API to produce API tests, page objects and locators, using Playwright + TypeScript and Axios, saving code and execution logs, running tests at runtime via child processes and automatic retries. A self‑healing loop sends failure logs back to the model, gets patched code and reexecutes; demos show many tests auto‑fixed but overall success after retries sits around 10–50% depending on instability. Costs stay low in examples (cent range), but reliability and edge cases require human oversight.
AI-written summary · about
chapters (18)
- 0:00GenAI in test automation
- 3:10QA roles and automation
- 6:20GenAI benefits for automation
- 7:04Prompt engineering basics
- 13:31GenAI as junior QA
- 22:08Prompts for automation tasks
- 22:30GenAI in test pipelines
- 32:29Demo setup
- 38:10ChatGPT API integration
- 40:19Framework config and execution
- 51:58Test code with ChatGPT
- 53:21Self-healing failed tests
- 1:00:12Self-healing mechanisms
- 1:07:25Demo with superhero API
- 1:12:29Demo with pet store API
- 1:17:36Codebase overview
- 1:17:57GenAI limitations in automation
- 1:22:36Q&A
related talks
01h 16mAutomation of the Automation using ChatGPT
26mAI and the Future of Testing
48mTestGPT: Generative AI (AutoGPT, ChatGPT & VisualGPT)
35mApp Testing with AI Tools – Make it Easy
37mAndreea Cosariu: An AI Experiment: Building a Test Automation Framework Using Generative AI
16mCan Test Driven Development be speeded up with Generative AI?
36mUnleashing Generative AI: Improving Developer Productivity in SRE
32mHow to Make AI Work for You: A Deep Dive Guide into ChatGPT for IT Professionals