Welcome to Software Testing Unleashed, the podcast for testers developers and all software people who want to create best quality software. My name is Ritchie. I'm a software-quality coach author & keynote speaker And happy to bring you an episode live from The Hustler Conference in Budapest. Hustev is one of the greatest conferences I've ever been to. Top quality talks, amazing people and that unique mix of friendly and intense atmosphere you only find at truly great events!
If you have ever get a chance go there seriously – You will love it! My guest today is Michael Buchko. Michael's software-quality manager focuses on quality processes in project deliveries. We talked about leadership in remote teams, how to deal with people when they are not at your place. How you can support them building trust and gain growth in these teams? And now enjoy the episode! But it's my first time. Okay, but I feel Elho at home still!
Actually this is the second time that i was here like seven years ago So now it's like five times bigger, but the atmosphere is still the same. It's one of my favorite conferences I would say in Europe. Yeah and we are here because you have a very interesting topic... ...I checked the abstract upfront who fits into the podcast then as well were doing test leadership at the role of testers And so on In the remote world.
That's very important Because today in my speech i get asked how to deal with remote people and there's no idea. But now we have to answer here! So that was the reason for my talk, so basically I started with when there was a Covid pandemic in Poland. There were no remote working before. They switched. all us as managers and leaders just go on remote work And figured it out. Now many companies are moving back into hybrid model People.
that hybrid is still remote because one of the team members is visiting office on a different day for example and The same practice is still applied. so And sometimes I'm still pushing some AI related stuff there. Yeah, Because the AI is One of their remote tools that we are using and Sometimes it's easy to use even by multiple persons during the same time like Multiple team members are sharing a single prompt between each other and they add to it, so that still needs to work remotely.
What were the problems when we worked remote? For me the biggest challenge was before Covid I was mostly working in... UK based companies and the practice that I was learning there, it's managing by walking. So we were like visiting our team members' desks speaking with them looking at their monitor... Oh! That is interesting. what you are doing. And basically on a remote work doesn't exist. so You have bigger challenge to build trust because you're not sitting in one desk.
There is a bigger challenge because in the office, meetings were limited by availability of meeting rooms. And that time you are working and now when you're remote then if your calendar's free I can push a meeting inside it... ...and suddenly you don't have time for team! Because i'm spending like... thirty six hours a week on the meetings. And then my team says that they had like hundreds of topics for me but never got time to head and catch up. so I'm trying show people what they need.
build this trust. That's one thing, it is one value. The second value i am seeing is advocating. So in past was really easy meet your boss or higher management in the kitchen, by a coffee table and then you could brag about what wonderful things have been done. And now with remote work there is no such possibility. so I'm trying to show people that basically team leader are the bridge between management and teams.
The only person who can make visibility of their work being done Is this leader To try send an email once every half year During the last half a year, that are achievements of my team. If their management doesn't have time to read through it... ...that's not problem. but if somebody sees and says oh! That is interesting. I thought someone else did this. And then suddenly shows up there was my team doing so Not different teams.
So i got an example where one component reduced number of open defects by about eighty percent. Management thought the reason was that the component owner change the priorities. But it turns out that I got like a quality manager in my team and asked him, can you ask all the component teams to work more on their maintenance? And then he achieved those results... He convinced the Component Owner to change his priorities.
It wasn't like the proactive idea of the Components Owner To lower the pace for features and start fixing issues. asked from the outside, trying to implement that. And then I could send an email... ...to show that Mateusz for my team did it and suddenly three or four persons in the leadership recognized him as a totally different guy…. …from a completely different team who was affecting their outcome of this component! Yeah so you celebrate your own wins and show them through management?
As you said by e-mail is there good way to communicate with the manager? I do it through an email because i feel like this is the only way of communication that's asynchronous and state. Like, if somebody sends you a message on Slack... ...you need to respond so they can't wait for months until I read them? And in the email, I feel it's not so impacting.
So even if i get a thank you half year later for that email... ...I can still go back to my team and said For one hundred percent this person read that e-mail.. ..and said Thank You! Yeah yeah If I write on Slack? I have no idea. because.... And do we need to find our manager who is on Slack?! That also the problem! Okay then thats very good. to celebrate their own wins. What else does he has?? in your toolbox for the remote teams?
For what else, so I think...for me the most important part is trust. So i work with my team and try to get their trust at every stage. .So I got like a meeting spaces on my calendars that we have meetings being organized like booked in my calendar, and it's a meeting when I arrive. And then ask how can i help you?
So basically the team member is covering all of our topics that we need to discuss or he or she can cancel their meetings Or they could say let us speak about something totally not related to work. Maybe somebody wants to change career path. So, I try to do it in a way not to lose the team members.
If somebody from my team says that they think about changing their career path... ...I can help them find mentors inside our company and then meanwhile reach out to leadership saying this person thinks of changing roles. If there is an opportunity on our company side, let's try to make that possible. If you don't have the trust, they will never tell you.
That's my plan that I'm switching to a scrum master next year and i don't care if there would still be scrum masters in our company... ...I am switching! Yeah yeah yeah.. I think trust is very important part when thinking of remote teams too because When Covid starts A lot of managers did not have trusted their people who work while sitting at home. And so now we see okay it often works more than in office. How can we gain trust to the management that they trust us? Do you have any tips for...
What I'm trying is just a transparent communication. So, i am showing them what we tried in like those half-a year or quarters and even saying about failures. so im saying We Tried to achieve this, but we failed. I'm not saying why we failed. maybe if somebody asks then we can discuss it. But i want them also to be like... Not only see the successes but see their work. That's my main focus. So even If We invested The time, we invest in the Work To try to do It? But we failed!
I still wanted to Be visible for the management that during This half a year those topics, the ones that we achieved. Those are the one's that were failed because sometimes their initiative might fail but you can get interest from them management and then maybe they could give somebody to help. So my problem was I had a task for my direct manager To review software development process of R&D department for the last three years. But they never produced a process, so it was always postponed!
And there were moments when I said that... ...I tried to ask this list of people about what the processes would be available because I got a target to review them but nobody gave me an answer. and then one of the department leaders says he can build workgroups in order design the process if you will help me to coordinate them.
And I'm running the coordination of five work groups in a different department, they are designing the process with me and then i'll review formal approval for that process to be used at this department. If not sending an email saying it was my failure over these last two years because we didn't deliver those processes nobody would give us priority or time to coordinate. So it gets you the result.
I think there are two very important things in that, The first is to openness and also communicate the failures because often we just show successes And other thing happens there To be opened up. so this was a problem On another side bonus track the problem to solve, get into a solution. Exactly and if you don't do the blaming then nobody knows what it failed. so... It's like transparent that this element fails but not because of them or me. I try to do it and fail.
If anybody can help achieve I can retry in the next half a year, maybe then i will achieve it. And if its priority issue somebody would respond and we said that they could give you someone to help. You mentioned that remote work together is changing because of AI stuff. now? What's your impact there? what do you mean? It's very complicated, I would say. So basically in my company we are a very limited usage of the AI.
The only thing that is approved to use is the Gemini because we're on Google office Stuck so you can use Gemini but mostly connected with documents. But basically with my team we observe what is happening on the market, that AI is growing. So for example me and two members of my team are running a side experiment in our private machines. We're using different models to research whether it's possible or not.
So I'm running an experiment, for example on developing a mobile application only with AI testing it. The project management being done by AI. I think the infrastructure is like twenty different agent personas and everybody has their own scope of responsibility. Um...I try to somehow connect during my talk about this next month in Milan at Testing United. so i want show everyone What did I achieve and what was the infrastructure?
And it's very interesting, like the pace of growth in AI is very interesting because there... ...I think they called for papers. that conference was half a year ago. Yeah! Then i got accepted.. ..and I've written already seven different versions of my presentation. And I'm afraid because the conference is somewhere around mid-November and there is a new model being released at the end of October.
And if I have time to switch, what the leaks are saying how bigger it's the Google Model then probably i will start my presentation from scratch. but yeah its mind blowing. Today, during the conference I was sitting in a speaker's room and... ...I remembered some blockade that i had on my project. And I asked the model about can you write me a solution to like a single sentence problem? That this infrastructure of AI agent could solve?
when.. ..i'm going back home then we'll run your command or instruction there on the models and basically, the model returned to me with instruction of one hundred pages generated in two minutes target at infrastructure, model infrastructure that I wrote. so when i try to read the document and have no idea what's there. And an interesting thing is estimate of delivery on this one hundred pages it four hours. It's growing really fast.
You mentioned before that the whole trust thing with them, what do you think when we use AI more in our product? The management says okay why don't we have now testers and developers there? We can reduce them halfway. What are your thoughts or what can be done to promote ourselves In this direction so as not to get rid of our job. I would think about it. That's why I'm convincing my team to experiment with AI.
So, my idea is that AI models should be something that a tester or any person doesn't matter. the role can have like you have...I don't know..like knife in your belt? Or axe in your backpack? so i would love everybody being able present their benefits. if I'm as a developer enhanced with AI, not in context if AI is better replacing me. And that's what i am thinking about. like how can i promote some ideas? That maybe AI could implement for me? but... I created.
Yeah, yeah and you review them now because of that's the issue of the trust. so in my experiment that i'm running i'm giving one hundred percent trust to the AI. So basically how i am trying do it is ask the AI design a prototype. then i asked him doesn't don't read it? Then I ask the models or agents to implement the prototype. then i'm doing like a human experience testing of the prototype and if im happy about the prototype, than I asked it make into production level.
If Im not happy with the prototype ,then write to the agent. this was the prototype...I don't like Build a new prototype, but then in the second try or third Try. I don't ask for an instruction. Yeah, I asked for a new implementation until we reach Working yeah prototype and then I ask it to document that prototype that I like And That's a huge change that took me few months. trust AI to this level?
Yes, and I think that's their challenge that exists In front of all the technical people because I see the management that they are very optimistic. Yeah, and i was reviewing some of their tools for AI automation And then I seen it from The background of the tool That you have like a fancy UI and direct API call to a model so You can do the same output asking them all itself directly without that Fancy UI and you don't have to pay five hundred thousand dollars for the tool.
But management has so much trust that they do not look at it in this critical way, which is a benefit of us as technical people... ...that we can critically judge if that tool could replace me or if its just a fancy UI with an engine behind them is inside the model. I think one of recent presentations by OpenAI, when they killed... ...I think they killed a API access that was available for their models and you now cannot directly use it as a microservice in your own.
startup and the guy when they were presenting that, he said there's a different pricing model. And all those capabilities will be built by OpenAI not by third-party tools reusing their API. Then he said openly we are right now killing over one hundred fifty thousand startups By closing that door. But we need to do it To make open AI profitable. You buy a tool for AI test automation For half million And then open AI with the new model says this API calls are closed.
Yeah, and your tool is useless. The only way how to predict it is to ask the technical people deep dive in that a I should buy it or not? Not like marketing guys sales speeches on all that stuff. It has to be done by them. Yeah I think don't buy this test automation tool, it's not like i'm defending myself to keep the job. Yeah! I am feeling that these tools are not a solution for us and thats the trust we need from management.
because then... We just use those tools which you dont want And thats recipe of disaster! That is great. Michel thankyou very much for your insights here about this topic. chatting later at the birthday party too and have some drink together. Yeah, of course! And yeah I appreciate that you were here part of this show... ...and i wish your a nice conference and hope to see you soon in other conferences. Thank You! Enjoy the Conference! Bye!