I have been working on reducing manual efforts on Manual testing and bringing in automation testing for a retail domain. The problem is the selectors keeps changing on updates on retail website (mostly every two weeks). Hence, this is mundane task as I need to make changes all the time. Can I Implement Agentic AI into Test Automation. If yes, How can I do so? (Please provide necessary document links, So I’ll refer on to it).
Can my agent conduct testing on the site, with needed info given?
Well, I am open to suggestion. Appreciate your support.
You can definitely reduce this kind of maintenance, but I would frame it carefully: Agentic AI can help in test automation, but it should support your framework, not fully replace good automation design.
In your case, since the retail site changes frequently and selectors break often, the first thing to improve is the stability of the automation layer. Before going fully into AI, try these:
Use stable anchors instead of relying only on dynamic selectors
Build reusable object repository/components
Add proper exception handling , will help you to debug
Agentic AI can help you
detects anomalies, flags flaky tests, and identifies redundant test cases from large-scale test executions
Generating test cases from requirements
Self-healing locator strategy
Creating test data
Analyzing failed runs and proposing fixes
I build AI agent to help Ui Automation. Please find the below link:
I went through the Screen play documentation and conducted a PoC too. It worked as expected.
I had already used healing selectors before (Semantic Selectors) while developing the cases keeping frequent changes in mind. But, In production it started to fail due to licensing issue, If feel there is a license limitation yet to figure it out.
For Screen play, what is the particular license that I need to look into ?
Well, I’ll look into this documentation and try to implement a PoC. If things are working expected. will let you know here.
Appreciate your support and quick response!!!
