Testival Meetup #73, Zagreb, Croatia

Estimated reading time is two minutes.

Meetup

This was a smaller meetup. About twenty people attended. During the usual introductions at the beginning of the meetup, I have noticed that people have arrived in groups. There were three groups of three people from the same company. (I didn’t catch how many people were from the host company.)

There was one regular talk.

• Oleksiy Slavutskyy, Testing of AI-driven Applications: how to not to fail when testing products with AI under the hood.

During the talk I learned about several tools I was not aware of (DeepEval, Ragas). I have also heard about small language models (SLMs). (I was only aware of large language models.)

I’m still as confused about testing AI-driven applications as I was before the talk. I did learn about a few interesting tools.

The most interesting question was what’s the biggest difference between testing normal applications and AI-driven applications. Oleksiy said there’s no test or fail. It’s always in between. For me that feels like we’re in Isaac Asimov’s Robot series. In the future, we might be more psychologists than testers.

There were a lot of questions after the talk. I might be subjective, but I felt other people were as confused with testing AI-driving applications as I was.

There were two lightning talks.

• Karlo Šmid talked about Livebook.
• I talked about a patch I had merged earlier that day. After so much talking about AI, the patch looked really insignificant.

After the meetup, somebody mentioned Jules. Another tool to take a look at.

Future

We’re looking for companies that will host future Testival meetups. We’re also looking for speakers. If you’re interested in hosting or speaking, contact Testival.

Links

• Testival Meetup #73 at testival.eu
• Testival Meetup #73 at meetup.com
• Testival Meetup #73 at github.com

Feedback

Thank you for reading. If you want to stay in touch please use the feed or send me an email.