What AI-powered testing tools have you found to be most impactful in your work?
I’m curious to learn about your experiences, how these tools have influenced your testing process, improved efficiency, or brought new challenges.
2 Likes
We are not using as such any specifc ai powered testing tool, we use chatgpt or claude for day to day task.
However in my last organization i have used katalon for e2e automation of project and my experience pretty much good with that tool. There were some issues while handling some complex scenarios especially those with dynamic xpath or those scenarios in which data appear for just few seconds like toast notification. W
Their support team was co-operative, so even though we faced issues but with their help and support it was resolved.
3 Likes
Thanks for your detailed answer @ujjwal.singh.
May I ask what type of projects you automated with Katalon?
And when you say for day to day tasks you use AI tools like ChatGPT or Claude, you meant for writing test cases, debugging..?
1 Like
It was an IOT based project. We had received a device from the client side, which was connected with the mobile application through Bluetooth, and in the mobile application, we fetched the reading from the device at regular intervals of 30 seconds and that data was sent to the backend and then the same data was visible on the admin CRM portal. So, we developed the mobile application, CRM portal, and web version of the mobile application.
Our goal was to automate E2E for all three applications, We were able to automate the web application up to 80% functionality however, in mobile, we faced a major issue while handling the animation. Their software wasn’t able to automate it. we tried to write a script manually for that part, but that also didn’t help much., Even their support and engineering team also tried to resolve the issue, but we couldn’t resolve the issue to the full extent.
One issue was that when the mobile application architecture was developed, the team hadn’t thought about automation. However, as the application features grew and it became complex to test, we tried to automate using Katalon but couldn’t succeed.
This was one of the major reasons we rejected the adoption of Katalon at the organization, as we were not completely satisfied with their mobile automation part considering the cost of their enterprise edition.
For second part, yes we use the chatgpt and other gen ai tools for test cases, debugging , scripts and for other use cases.
2 Likes
Github Copilot -
Saves time, frees you from low level coding work.
2 Likes
We also use Github Copilot.
This is a great support for programming. You can also use it to generate unit tests, for example. However, I have found that the more complex the object you want to test is, the less useful the test cases are.
2 Likes
I use ChatGPT as a general sparring partner, and previously to help with coding. I just got access to GitHub Copilot though, and I’d definitely say that’s more convenient for the latter, as it’s integrated into the IDE. Less time spent copying and pasting, and it’s a bit like predictive text for coding. Speeds things up, but get’s it wrong a fair amount of the time too.
2 Likes
We have used chat gpt for asking how should we write an Test Specification Document that fits in the scope of the Functional Safety standar ISO26262.
The chat GPT literally wrote with details all the sections we should include in the document. How descriptions should be done and how we should link with Requirements for have correct trazability.
Also, he told us how we should document the required reviews of this Specification.
The document at the end was accepted by the client, so I guess it was done. We had literally no idea about the ISO26262 neccesary insights. Chat GPT cleared all this for us.
2 Likes
Here is the list of tools i have come cross and others around same. But to really to know and understand which is best is really also depends ur context, and which Testing activity u want to use AI for.
1 Like