DATA_FAIR Conference

This week I have attended the DATA_FAIR, data engineering and data science conference here in Ljubljana. I regret, that I could not participate fully, but my body decided to make the first day of menstruation even stronger than usual. So a lot of resting and some sleeping happened.

(I also really do not have luck with timing, as I also got it during the IndieWeb camp in Nuremberg, except that time it was more manageable)

Still, there is one thing that I need to talk about, that I think they did really well. In difference to the normal tech conference. The majority of the people there were presenting as women. And when I say majority, I mean very noticeable majority, not the 'one has to count to be sure' majority.

I am not 100% sure how they did this. But there were some indications.

The first would be the roster of people presenting. If you look over their pictures and name on the speaker section of the conference site, you will see one thing - all of them are women. Some people reading this are probably not familiar with the local names here, but all but one are quite common regional female names.

For anybody that wants to explore the personal sites of the local (to me) female tech, I have tried to find the websites of every speaker at the conference (most unsuccessfully). You can find the list below.

• Tuuli Pöllänen
• Alja Isaković
• Božidara Cvetković
• Maja Ferle

The second would be the on-site childcare The public kindergartens have the ending time. While I am not sure what the current is - it was before 4 o'clock when I was there. Still, if one needs to pick up the children, then they can not stay for the entire time.

Before the conference even started, I talked to another participant, who took advantage of that. It was an adventure for the child, and she could stay to the end.

It is a good question - if you want to bring more people of a certain group, it is nice to ask them what is stopping them. I know that a lot of women answered children-related tasks, so it was nice that organisers listen.

(When it comes to local conferences, this is not the first time I saw that. It is becoming more and more frequent. I did not really see that in many tech conferences abroad.)

I for example know, that at least one foreign queer community lists the availability of the sign interpreter at their events. I know queer community here that lists the languages that the organisers speak and what language can be expected on the meeting. If any of you are thinking that English is accessible default option, I will come after you with the butter knife. This week, I have requested to join a 18+ Discord fandom server.

All of these are examples of how to make people more comfortable and more likely to participate. They are just different people.

The third one is, that there were no diversity talks. Honestly, the only time diversity was mentioned was, when one of the principles of non-biased AI was having the diverse data and diverse teams. It was less than a minute in the entire day.

In at least some conferences (for example PyCons), I have noticed that it is popular to have a diversity talks or even tracks. And I am not going to say, these can be interesting. But why it is basically always a woman than presents these topics?

I am glad, that here this entire problem was side-stepped. We could instead listen to the women presenting their projects, from basket analysis dashboards for a supermarket in the neighbouring country, the supply chain optimisations for a shoes company, the picture selection algorithms for an advertisement, prototyping tool for data products or data infrastructure in the company responsible for electrical grid.

And the not hard-core text problems were bias in generative AI and copyright legislation. Both very important and very relevant to the data engineering and science topic of the conference.

It is hard to explain, but these choice of who and what to present did change the vibe of the conference. It would be possible to point to the specific points, but I don't think they would really be able to bring forth that vibe. Or maybe outlook would be a better word?

I guess I can give the contrast example. In one of the presentation, they presented the design for the AI phone support. The model started with how the user will have a choice to wait for the human or have AI help them. And then what data sources will be used and the whole loops.

It was a generally polite technology.

And then there was a question from the audience. They had implemented the AI phone support by making AI pretend to be a sales person. But the AI had the problem of using any kind of inhale or short pause as the end of talking, even if the person was on the middle of the sentence.

That I would describe as the non-polite technology.

(Politeness of the technology is an actual scientific term that we talked about at the Ethical Computing class. The professor has a website, in case the concepts similar to that are of interest.)

It was also a conference, where they told us to be annoying to our employers, where everybody on the stage but also the small thing that I noticed outside, had the constant support and point out the great things other people did. I did not expect the substantiality being mentioned multiple times, even though it was not the main topic of any talk. I was surprised of how there was a speaker in the round table, that kept pushing back on the need for generative AI and pointed out how much the non-generative AI had also helped us. Where they told us to question the need for the analytics and metrics and instead talk to people.

(Which was even more interesting, because just an evening before somebody was convincing us to not use anecdotes but instead metrics, since these are better. :) )

There were probably many other things, that I did not notice. I have to say, that the organisers did a great job. And I think I can continue to learn from this example.