Episode 105

Igor Polyansky on AI is transforming hiring by streamlining processes and enhancing decision-making

TIM: We are live on the Objective Hiring Show with Igor. Igor, welcome. Thank you so much for joining us today.

IGOR: Hello everyone. And thank you for inviting me.

TIM: It's our pleasure to have you. And I think a great place to start would be just to hear a little bit more about yourself. Who is Igor? Who are we talking to today?

IGOR: Yeah, I'm Igor, a data analyst. Yeah. In the past, the current data analytics leader in the Muse group made some products for musicians, from Ultimate Guitar to Hell Leonard. My role is a typical data AI and analytics leader. So I build in the teams to create appropriate decision-making based on the data inside the company, building some tools based on AI and so on and so forth. As my side activities, I also listen to some podcasts like this. So I'm working on my personal brand, and I own the analytical consulting agency. So we provide services for the companies to build in appropriate data teams to monetize and implement AI-driven drive services for the companies. The company is FaktLab. You can find it on LinkedIn.

TIM: An excellent intro. And yeah, what an interesting place to start now, actually, if I think about it, because you're at the cutting edge of a lot of this stuff. Do you think AI is overhyped? Do you think it's the real deal? Is it really changing things as much as people feel it is?

IGOR: I think it's quite common for any new technology, and yes, it's overhyped right now. If we can make the difference with AI also, yes, but as usual it depends on how you use this technology. So I see a lot of cases I'm involved in where huge companies are using AI right now. And earn millions of dollars because of this impact. But at the same time, there are a lot of companies who are struggling with implementing AI because of data quality problems. In an organization in terms of the structure, infrastructure, and so on. It depends on your company and your maturity, and how are we actually ready to implement such things?

TIM: One interesting lens I was thinking about earlier today was, are we currently in a situation where AI is on average overused or underused? And what I mean by that is that they're more things that people are just trying to whack with the A.I. hammer that is never going to be a nail. It's not going to go into the piece of wood, or they're actually way more opportunities that are almost untouched that haven't been sold but could be solved with AI at the moment.

IGOR: Yeah, a hundred percent. I see the same on the market. There are high-tech companies with energetic leaders who implement AI everywhere. But at the end of the day, AI is not. Always the best solution for any case for automating any process. For a typical implementation of AI stuff, you start with simpler things. For example, you want to automate some processes. What should you do? Yeah, first you need to digitize it. Maybe you can automate without AI based on simple rules. And then implementing AI to make the final difference. But in these companies, they deep dive into AI in the first steps. And they struggle with implementation because they didn't pass these first steps. on the opposite. The companies that are not involved in AI enough for 2025, maybe it also depends on the leaders. And if leaders are not so driven by new technologies, maybe they're skeptical about AI roles. They believe that AI replaces any professional, and it's stupid. And so on and so forth, they are under the general market in implementation. But I believe the situation will change in the next year because. Yeah, AI is a common sense right now; the companies that implemented it are just more efficient at the end of the day. So they will compete better, and companies without AI will lose this competition in the next year. So they will quickly realize that it's a vital step to develop a company. And we will see AIs in most of the companies in the next five to seven years, I believe.

TIM: I would have guessed and would have felt like because at least with the L. M. changes and improvements we've seen in the last couple of years because it's been so profound. I would have thought there can't be that many leaders of companies who are still like, Oh, AI is all bullshit. I've never even used ChatGPT. It's all overhyped because it's so big. It's not like it's a small incremental change. It's like a boom, profound change. I would have hoped there wouldn't be that many completely head-in-the-sand leaders. If someone was listening to this and they feel like the leader of their business is maybe not as open-minded. about AI as they should be. Any ideas of how they could get them on board? Would it almost be helping that leader use AI in their own life? Hey, Mr. Business Leader, let me help you install ChatGPT on your phone or Claude. I can help you figure out your recipes on the weekend, or I don't know, whatever else; any ideas on how to get someone thinking about adopting AI when they're against it?

IGOR: I think I met the same challenge about 10 years ago when I started working in analytics, and it was not hyped yet. And it was not common sense for the market that we use data-driven insights and data-driven decision-making, and it improves the business. And it was quite the same. Business leaders with It's old-fashioned mindsets that, oh, I don't need it because I fully understand what I should do. I've been working here for 20 years, for example, and I don't need any advice. But now it's common sense for most of the companies to invest in AI data and BI, and it helps us day to make decisions better. What did dialogue 10 years ago look like? You just need to find someone in the company who can buy in for this new technology and make a business case with this leader together. As you, you show real business results. AI, you can involve more people in using it. You will come to the business leader, not with, Oh, you, AI cool technology, let's use it. No, you show them a case. We implemented AI into this process. It automated this process, and it saved us. 1,000,000 in the costs, maybe; it makes sense to implement the same for other processes. When you speak with the leaders on the business language, it helps most of the time. Of course it's a more long-term conversation, but in the end, the language of facts works in a truly data-driven world.

TIM: Yeah. Money talks. You mentioned, yeah, maybe 10 years ago, facing a similar experience of having maybe some claims of let's use data to make decisions falling on deaf ears. I personally feel like. That's been the case in recruitment and hiring for quite a long time, in the sense that I think most people, when they're hiring, use, as the majority of their decision, their gut feel or their intuition rather than any metrics. And even if there was some data about a candidate, to be honest, most people would ignore it if they just didn't like the candidate. If there was just some reason they thought, I don't know, I don't think they're the right person for the job. I can't even articulate why I just don't like them. So will we get to a point in the next few years where hiring is largely driven by numbers and metrics rather than feelings and emotions? Do you think?

IGOR: Yeah, I believe it should become true because now gut feeling or gut instinct follows many decisions in the hiring. As a hiring manager, I fully understand this bias because all managers have had at least one case, or even more, when they followed their gut instinct, and it was a successful hiring. But at the same time, it's a selection problem. When we forget the cases where we followed our gut feeling, but it wasn't successful. And in the end, you believe that you have a lot of experience and your gut feeling works. For me, it's about the balance in the end, because gut feeling truly matters. I personally faced a lot of cases where it worked because of the rich experience of a specific person, a specific manager hiring, or some business challenges. But, of course, we can't ignore the real data, and the final decision should be data-backed. And intuition should be just one component for such decision-making.

TIM: What I've sometimes thought about was having, let's say, an objective hiring process where you think about the things you're trying to evaluate in the candidate, which could be soft skills, technical skills, or amount of experience. You have some kind of matrix where you then score candidates along it. And then people would often say, Yeah, but there's always this kind of like black box, intuitive gut feel element, which I said, Fine, no worries. Couldn't we just have that final gut feel bit as just an extra metric? It's, I'm going to say, I don't know, my weightings are 80 percent logical and numerical, and then I'm going to have 20 percent gut feel. I don't have to tell you how I came up with this. I'm just going to rate the candidate based on my pure gut. I'm going to give them a 15 out of 20. That's worth 20%. Because then at least we would have called it out and said it's still a factor. We agree. It's important. I don't quite know why or how, but at least we can incorporate it and still make it a number like everything else. What do you think of that pro that kind of approach? Could that potentially work?

IGOR: I love it, actually. It's a typical data-driven approach. I can make an example not from hiring but from product management. In product management, we work with hypotheses a lot. What is a hypothesis? It's our prediction of potential impact if we change something; for example, I will recolor this button, and it will increase our conversion by 10 points. How do we work with it? We use some frameworks for decision-making, and one part of these frameworks is usually confidence. What does it mean? It's our assessment of the risk of making such a decision, and it consists of several parts. For example, you have your gut feeling, and it's anecdotal approval and justification of this hypothesis. It's about 5 percent of our confidence. What's next? We can deep dive more and invest more into the research. For example, we can ask experts about this hypothesis. It improves our confidence to 10%. Then we can do additional research, conduct UX interviews, collect more data, and improve our confidence step by step. In the end, of course, the more objective. More objective approval of our confidence will be a more data-driven approach. So the final validation will be, of course, MVP, where we launch the actual hypothesis and see real business results. But we can have the same in hiring, of course. The hiring manager, HR, and other people who were involved in the process can contribute with their gut feeling about this candidate. But it can slightly improve our confidence in this candidate. But the main part of this confidence should be based on the objective parts like skill assessment, experience assessment, reference check, and so on and so forth. I really love such frameworks, and I think it could help to align expectations inside the company. Because the typical problem of hiring I see is that the hiring process candidate by candidate is different, unfortunately, because it's a people factor. It's different interviews; for example, the technical stage for the analyst position in my company could be done by several analysts just to distribute the workload. And of course they have slightly different approaches to how to conduct these interviews. But having this magic of competence, this framework gives us the main foundation for how we make the decisions. And in the end, I fully support such things in hiring as well.

TIM: Yeah, it's so hard to control and make something consistent and truly structured if it's done by humans at the end of the day, even with, as you say, those different interviewers having the same set of questions, the same kind of scoring rubric, or what have you. They're still going to interpret the candidates answers differently; each interview is still going to be different because each question is going to be answered for a different length of time. And you're going to drill in to certain details and not in others. It's just, there's just almost infinitely many parts it could take. Okay. Which is why I feel like AI would be such a great tool in hiring because it would have the ability to structure things in a much more consistent way and at least remove the individual human interviewer bias. Maybe it's got other problems, but I feel like there's a lot of upside. How do you think about AI and hiring? Do you, are you bullish on it? Do you feel like this? Yeah. Where do you think it's going to go?

IGOR: I think we can involve AI in each stage of hiring. Honestly speaking, talking from me right now, I am still reviewing CVs myself. I can't fully rely on the AI yet, but I do believe that by the end of 2025 we'll have AI tools handling initial candidate assessment. It could help summarize key points and offer recommendations to the hiring manager or HR through the interview. You should ask about this specific project, or it's not enough information about this specific skill. So let's deep dive into this. It could highlight even the key points. In the CV, just to screen it better when you were reviewing the CV, you usually have just one or two minutes. So an AI-aggregated summary could highlight the most important recommendation to HR or the hiring managers. One interesting scene, even in 2025, when the internet is flooded with. Advice like your CV should be just one or two pages; your CV should be shorter. I still see some candidates with massive detailed CVs with multiple pages. And honestly, sometimes those are the most interesting people for me, but I still have one, two minutes, maybe in the future. As AI during assessment becomes standard, CVs will actually get longer, which I expect. So you can pack your CV with a lot of detailed experience, a portfolio, maybe some projects and skills, and tools you worked with, and AI could extract. What is the most important tip for your specific position and help hiring manager with such a summary, and it could make hiring much more efficient because of the initial assessment? I would already be much deeper in my agency; we have several projects to test such approaches in the companies. And it looks like this. You ask the candidate not to shorten the CV, make it as detailed as possible, and then you're trying to auto-assess it with AI. AI gives you really interesting results on the pilot stages. So I believe it will help; maybe the second part is some help and an assessment with the real interviews, people to people, because most of the time. As a hiring manager, when you interview people, you make some notes in parallel; it decreases your focus on the candidate in the real conversation because you're scared to forget about some important points. You're scared to forget about some important questions. but. If you have an AI-based tool to record, then transcribe and summarize your conversation, you don't need to worry about it. And on the other side, you can have not just a tutor but maybe an AI partner, AI partners that remind you about important questions, about something you should deep dive into as a candidate. And you can fully focus on the conversation here and now. And make it more detailed, friendly, and people-oriented. I believe that we have a huge room for improvement with AI in hiring.

TIM: Yeah, I couldn't agree more. And the first point you made around the length of the CVs is a really interesting one, because I feel like one of the fundamental issues with hiring at the screening stage, especially, is the data set that each side has to decide is so poor, like a company has a. Yeah, a two-page CV, maybe, and maybe some questions from an application form sometimes. And then the candidate has a job description, and it's a very weak set of data to make any matching on. So if, yeah, this new world where you have an elaborated CV that's going to a lot of detail, and then almost you're like an AI sitting on top of that to summarize and extract, that could be really interesting. Because you can always go from detailed to summarized. You can't go from summarized to detailed. So I think that would be a big improvement. And yeah, there shouldn't be any limit on how long an AI CV could screen because it's just a few more tokens.

IGOR: Yeah, a

TIM: one,

IGOR: And there is not another problem with datasets, because if your previous hiring was subjective in some way, the model you developed based on this dataset will also be subjective. Of course, data is everything, and maybe one of the biggest challenges right now is implementing such things, which will be cleaning the data and making this follow the rules you want to follow in the end product and AI product. So if you want it to be subjective, okay. But if you go to objective data-driven decision-making, it will need additional investments. So maybe you've seen some positions like AI professionals for HR, and those people have already made some progress in this area. So they fine-tune their data sets that the company already had to improve it and to make objective models, even based on the previous history.

TIM: So this is based on the previous hires that the company had made, like trying to search for other similar candidates and trying to score other similar candidates based on who had successfully been hired. Is that the approach that they're taking? Do you know?

IGOR: Yeah. It's not even about success, but about how the candidate will fit with the Yeah. Maybe you can summarize it with success, but in the end we can evaluate. Were there some objective signals that were ignored in the hiring process? In the end, an unsuccessful case. For example, a candidate had an average reference check, but you hired him for the high position with high expectations, and the candidate failed. It could be a risk; it could be a rational risk, but sometimes it was not a rational risk. It was just a gut feeling, and we need to reduce such factors with cleaner data sets.

TIM: You'd mentioned in passing that with your consultancy, you'd done some projects in this space. Is that right?

IGOR: Yeah, I'm currently on URL also in this.

TIM: Excellent. If you could chat about one of those, that would be great. Of course, without mentioning names or giving away any key details of your secret sauce. But yeah, I'd be really keen to hear more about your experience there.

IGOR: Yeah, I can add some details about what I've already mentioned. This is a project for a non-tech company. It's a factory basically in engineering, and they produced some metal products, and they had the challenge to implement AI into some semi-engineering, semi-IT jobs. And they didn't have a huge data set for this to learn from, too. To make a prediction of whether candidates will be successful or not. So what we decided to do is ask candidates to provide as much information as they have about candidate experience, share the portfolio, LinkedIn, your CV, cover letter, and anything you can share about yourself, and don't limit yourself with this one or two wagers. And up to this, we use several LLMs. To summarize this experience related to the specific position and highlight the most important areas. We need to deep dive in the interview first and red flags that would prevent us from, in the first steps of hiring on the pilot stage, having very good results because very few candidates in the end approach this approach this way. So they invest the time to make their CVs. Better and more detailed, but the ones who did it. They were more motivated at first, and secondly, they gave us this useful data set. And currently, the hiring manager and HR have a list of recommendations generated by this AI model. What should you ask about the candidate? This is an interesting case, and you could deep dive into this. And in context, that candidate will need the same experience in this position. Yeah, from the initial feedback and initial data, we improved the funnel in this hiring because we just dropped the applications with low details. Maybe it's not so fair, but it was requested in a job description. So candidates who didn't follow it, we just dropped this applications from first and we have a lot lower funnel with fewer candidates, but with a more qualitative. Feedback, and it works better in terms of conversions from stage to stage, and we already closed seven to ten roles using this tool.

TIM: It's an interesting project. And yeah, great to hear something that's happening like right now. With AI and hiring. So that's, yeah, really cool to hear. Yeah, I imagine that kind of approach would be a bit market dependent just because in the current market climate, for a lot of companies, they've got an excess of candidates. There's too many, almost like too many people applying. And so in that scenario, you can almost afford to ask a bit more, gather a bit more data to have a better prediction. But then now we might've had a 2021 market where it's like, Oh, there's no candidates anywhere in the world in tech and data. We have to just grab whoever we want. So it's almost like the, I've sometimes felt like in hiring, yeah, the approach you take and how much data and how accurate you can be is almost dependent upon what you can get away with in a sense. With the market conditions, is that fair to say, do you think?

IGOR: That's a good question. I don't have a final answer. Is it good or not, this candidate flute, when you have thousands of applications to your position? But my initial thought about it is that I don't see a real problem because AI is helping both sides in this road of AIs. has an advantage here in resources because employers could invest more into AI to improve initial assessment and drop non-targeted candidates. if candidates use AI to improve their CVs. Also use AI to analyze them. It's just the race between two AIs and whose AI is better. AI skills, I believe, are expected from candidates anyway. So if candidates use AI, it's a good flag for us. It's actually a plus, not a minus. Learn some new skills. Maybe the real problem is if the candidate lies or exaggerates his experience. In CVE using AI. But interviews, we still have them, and interviews will expose those gaps anyway, because it's much harder to fake expertise in real conversation. I know that some high-tech candidates are even using AI during interviews. But it's quite difficult right now. So there are only a few. and companies on the opposite have some methods and already have some tools to work with it. For example, neural networks that detect and analyze eye movements to show that the candidate is reading from the generated script. Or some models that are trying to identify AI-generated text. Don't forget about other stages. For example, reference checks help to uncover things that candidates might prefer to hide or look better on the page. Personally, I always conduct reference checks, and because of my network, I often know someone who has worked with the candidate before. So that's, insight is invaluable. So maybe the conclusion of this flooding on the market with many candidates could tend to an interesting trend, we might say, in the next years: faster hiring and faster resignations. If you understand that. The hiring process is not so reliable because of AI help; we will detect their real candidate skills in their first months of work. So if AI helps candidates look better on the page, it may just speed up hiring decisions. And then the resigning candidate also made it faster and aligned the whole process. I'm taking it into account, but maybe an AI just will compensate another AI, and we'll see just the balance. I dunno, it'll be interesting anyway.

TIM: It certainly feels, at least currently, like the explosion, the explosional improvement in LLMs, has broken the hiring process. Because I think candidates were very quick to adopt Chachipiti almost immediately. Companies, of course, it takes a little bit longer; they have to either get people like you to build them custom software, or they have to wait for the HR tech to catch up. And it's just, it's slower decision-making and all sorts of other reasons. So I feel like the companies are maybe slightly behind where the candidates are using AI in hiring anyway. So then, yeah, a lot of candidates, from what I could tell, are using ChatGPT to make their CVs look more like the job ad, which then, what I hear a lot is, Oh, I'm getting all these CVs, many of which look really good. And it's almost now, how do I choose from these 500 good-looking CVs, which of the, I don't know, five people I need to interview, because they all seem to look good? I wonder though, then, if we had a truth index about a CV, like how truthful is this CV, is that dropped now, if candidates are using AI to rewrite them or make them seem like the job ad more, what are you feeling? What are you seeing?

IGOR: Honestly, I don't see a real problem with making your CV better using AI if you don't have your experience. So it just will save the time of the recruiter and hiring manager if you. CV is better in terms of this visual noise and unnecessary skills and so forth. But for the other market, it's just the competition of different companies. Companies that implement such technologies faster will have an advantage. They will hire better candidates with lower costs in the end. And the companies that delay such decisions, they will lose in this competition. And someday the market just will balance everything. And such tools will be a must-have for any hiring process. So I don't believe it's something because you still have a real interview where you can assess everything you want to assess. Yeah. And you have all the tools already in place. And if you, the company, are massive hiring, you understand that. You rely a lot on this hiring in terms of business goals. You should invest more into such tools, taking into account the current market. That's pretty much it.

TIM: When you are interviewing candidates, I think of having just seen a post today from Anthropic, like Claude's owner, from their own careers page, so a pioneer in AI, and they had on their careers page on the job ad, it had a little disclaimer: We don't want you to use AI at all in this hiring process; we want to be able to evaluate your skills without the impact of AI, which, of course, you're going to use in your job, but we want to just evaluate you. And that was their philosophy as an AI company. And I personally don't know where I sit on this. Because I'm like, on one hand, why would you want to discourage someone from using something that they're going to have to use in their job? And if they didn't use it, they'd be ineffective. Why would you want to discourage that in the hiring process? But then what are you even evaluating that in the hiring process? If it's just a candidate who's looking at. Chachapiti has augmented their CV, has inflated things, and is maybe even in a live interview remotely, somehow getting hints from Chachapiti or Claude, like, who are you even dealing with now? Is this really the candidate, or is this candidate plus AI? Is it like a cyborg? What, where, yeah, where do you sit on this? What do you think?

IGOR: Actually, interesting question. I believe that such statements are more PR-based than the reality, because the reality is in 2025. The world is AI-based, and it's growing. If you don't use AI, you just lose time. But if you use AI for your HR process and hiring process, it doesn't mean that you fully exclude the people from this process. It just means that you make your operations more efficient and give. additional focus to more for better candidates for your role. So if you exclude 700 candidates from your funnel and keep just 300. assessed by AI. It gives your hiring managers an additional minute or two to review each application. So for me, it's about improving the quality, and it's even more about people in the end because we're just competing with AIs, and yeah, if some stages were made with AI, we should add AI on the opposite side to compensate. Then, but I don't believe that we will fully automate the hiring process with AI, excluding people, maybe only for some specific and mass positions. I believe it's possible. I saw some cases like hiring operators of customer support using only AI, but for more difficult and more people-based positions in IT. Currently it's quite impossible. So anyway, you need to check the feed with the manager. You need to check the feed with the company values and with the team. And it's quite important to make this with AI because AI has none of this. Empathy is needed to assess such parts of the coin. For me, it's not a real problem, and the companies that really don't use tools to improve their process, they just waste their time, maybe in the case of this PR statement, but for other companies, what are the most? I believe they will improve the process to improve the final impact from the candidate, and in the end, it's a win for the candidate and the company because if the candidate is in the right place for his skills and it's a good fit, of course, both sides are winning. The candidate is happier. Yeah, he performs better, and the company is growing better because he has these high-performing candidates.

TIM: I feel like AI could even. take care of. I feel like I could be used for the whole hiring process soon. You mentioned, for, let's say, more standardized, simplified roles. It already is, so fair enough. But even then, let's say, slightly wider, more difficult roles, where, as you say, there's like a fit with a team, a fit with a company, these kinds of things. Yeah, you're right. I don't really see how an AI could do the emotional element that a human could, but if our overall hiring accuracy is so low, I don't know. I don't know what it is in aggregate. People say maybe 70 to 80%. If you're a good hirer, you're getting it right that amount of the time. That's still a pretty big error percentage. So there's still some level of, like, overconfidence in our ability to judge people and make decisions. If we outsource the whole thing to AI, maybe we'd get an overall better accuracy score than that. We just have to sit back and let it do its thing and go. I kind of trust in this system that it's, on average, better than me. But maybe we're a long way from thinking like that. I don't know. What do you reckon?

IGOR: Yeah, I fully agree with you. When you speak with a person, you acquire many more signals. Then any two connected as a recorder of this conversation, because you have this emotional content; you see some nonverbal signals, maybe even that they're not clear for you, but you still have this gut feeling as a company of overall decision-making, how people actually sing, and how people make decisions. I don't believe that AI could make the same quality decisions as a person in the end using the limited data set of these unverbal signals. Maybe in some era we will implement some chips into our heads. To fix everything. All the emotions, your blood pressure, pulse, and so on. AI will be better in decision-making than people. But now it's quite impossible because any digital tool could be cheated. In the end of the day, because any digital tool contains some algorithm inside, one day this algorithm will be leaked on the internet, and people will understand how to cheat it and how to avoid any problems with it. But people are not the same. The algorithm is less structured than the tools, with the computer. So much harder to predict what question will be next and what will be your reaction as a hiring manager for a specific answer. As an HR person, you have a set of some tricky questions, some behavioral questions, and you have your own understanding of how to assess them. And it's quite difficult to cheat with them if you have enough experience with asking these questions. seeing the candidate's response and evaluating this response. So in the end, as I said multiple times before, it's about the balance. With our two worlds, AI and people, the synergy will make the difference.

TIM: I feel like one monkey in the room that we haven't really touched on is if, whether we use humans or AI to do the evaluation of a CV, a reference, an interview, or whatever it is. One big problem with hiring is that people aren't necessarily themselves during the hiring process. You do an interview, and it takes a really skilled interviewer and the right environment to help them relax, to actually have an authentic conversation where you are really seeing the true person. And so that would be an amazing unlock somehow if we could. Figure out the real person during the hiring process or maybe back to your point before it's just about collecting enough Existing data points about the person that you have then built up a more holistic picture Anyway, rather than relying purely on the interviews to do that.

IGOR: I understand what magic you mean when you speak of this skilled interviewer; you feel comfortable, and you feel less pressure, and it helps you to uncover your skills and your experience better. So thank you. Cool. I'm not sure about some; maybe some people say that I'm more comfortable with just listening to me and asking some questions than with the person. But I believe that people approach this. Creating this atmosphere of relaxation that is just a real conversation with two, two, two people. We are two professionals to each other, not too hard to evaluate you as a candidate. No. But to. Explain about our company better, talk about expectations, and then ask you how these expectations match with your experience. It's a much warmer discussion than just question, answer, question, answer with AI. I truly believe that candidates can. You can feel less stressed, and you will understand this person better in the end. But, of course, it depends on the seniority of the position you're asking for or looking for. For example, if you are looking for junior candidates, like interns or associates. Sometimes you want to apply even more pressure through the interview because you have no working history to evaluate. So you need to see how they will perform under real stress. because the first jobs might be the most stressful in any career, but for more senior candidates, you have this work interaction. It's additional information, and for you, what is more important is how they think, not just how they react under pressure. And more experience means higher expectations on the start. So that makes interview much more stressful by default. Even if you're asking a person to give some details about the job he was in five years ago, it's already stressful because it's quite difficult to remember. Some details for five five-year-old projects. So that's enough stress on the interview by default. So even if we can do it, it is less stressful; it will be good for both sides for the candidate to show his real experience and skills and for the company to evaluate the candidate better.

TIM: Less stress sounds pretty good to me. Igor, if you could ask our next guest a question about hiring or AI. What would you like to ask them?

IGOR: Actually, I will also ask about this concern of a fluid market. So we have mass job applications, auto-generated with AI, automated with some tools, and these AI tools make the market; they change the market. So it's, this makes it harder for strong candidates to stand out. I'd ask, do you see this as a problem, and how do you personally approach this? What is your strategy inside the company to tackle it?

TIM: A very important and very relevant question that I'm sure everyone will be keen to hear the answer to, Igor. It's been a great conversation today. We've covered a lot of ground, particularly around AI. It was great to hear especially about your current live projects in that space, because I think it's great to hear from people who are actually out there doing it. So thank you so much for sharing those insights with our audience today.

IGOR: Thank you. I'd love to share it. Interesting conversation. Related to AI and AI plus people, it's really interesting for me. So it was a pleasure.