Recruiting Future helps Talent Acquisition teams drive measurable value by developing strategic capability in Foresight, Influence, Talent, and Technology. In this episode, we’ll be focusing specifically on technology.
The fundamental shift from complex software user interfaces to conversational recruiting technology is happening right now, with profound implications for how we think about the ATS.
What are the implications for both recruiters and candidates when the recruiting system becomes an invisible part of the flow of communication?
My guest this week is Michal Nowak, SVP of Engineering at SmartRecruiters. In our conversation, Michal shares his insights on how conversational interfaces can transform recruiting from an admin-heavy profession to one focused on meaningful human connections and strategic thinking.
In the interview, we discuss:
• How AI has shifted development priorities
• Keeping up with the technology
• Working with TA professionals to spot use cases that don’t exist yet
• How AI can help recruiters build connections faster
• Moving to chat-based interfaces
• Using adaptive UI to deal with complexity
• Ethics, transparency, and trust
• What does the future look like?
Follow this podcast on Apple Podcasts.
Follow this podcast on Spotify.
00:00
Matt Alder
For decades, recruiting software has meant clicking buttons and navigating menus. Now, with the power of AI software, engineering teams are creating interfaces that you can simply have conversations with. To discover why this shift could completely transform how we think about recruiting technology. Just keep listening. Support for this podcast comes from Smart Recruiters. Are you looking to supercharge your hiring? Meet Winston, Smart Recruiter’s AI powered companion. I’ve had a demo of Winston. The capabilities are extremely powerful and it’s been crafted to elevate hiring to a whole new level. This AI sidekick goes beyond the usual assistant, handling all the time consuming admin work, so you can focus on connecting with top talent and making better hiring decisions. From screening candidates to scheduling interviews, Winston manages it all with AI precision, keeping the hiring process fast, smart and effective.
01:02
Matt Alder
Head over to smartrecruiters.com and see how Winston can deliver superhuman results.
01:09
Matt Alder
Hi, Michal, and welcome to the podcast.
01:12
Michal Nowak
Hi, Matt. Hello. Thanks for having me.
01:15
Matt Alder
It’s an absolute pleasure to have you on the show. Please, could you introduce yourself and tell everyone what you do?
01:21
Michal Nowak
Sure thing. Hey, my name is Michal Nowak. I’m the SVP of Engineering at Smart Recruiters. And in this role I oversee product development, infrastructure, developer experience and security functions in the company. And I joined in 2012. So, like, I’m one of the dinosaurs in the company. I was actually like engineer number six and I stayed since then. I’m based in Krakow, Poland.
01:50
Matt Alder
It’s just brilliant to have a software engineer on the show and talk to someone who’s actually building the stuff that we talk about day in, day out. So I’ve got some sort of burning questions for you. The first one is, you know, and as you say, you’ve been at Smart Recruiters a long time, so obviously technology has changed kind of since then. I would imagine you’re now building an AI first platform. How do you keep up with the technology? How do you make sure what you’re building is still relevant by the time it comes out? How does it all work?
02:20
Michal Nowak
It’s interesting, right? We’ve been here for over a decade, right? So we’ve been doing this since 2011 12. And that. That was when the first versions of smart computers came up and we started launching AI products in 2017. So, like 80 years ago. But I can say that were AI first back then. I think everything changed for us. Like early last year. We basically look at our product roadmap and we decided to cancel all of that. We message all of our customers and tell them hey, there will be no releases for the upcoming three months. We need to hit a pause and rethink what are we doing. And that was that pivotal moment for us where we basically thought, hey, are these features even relevant? Or will they be relevant in six months with all the technology advancements happening around us?
03:16
Michal Nowak
And that’s really where we changed and we reimagined the entire product roadmap. And that’s how we started this shift. And I think, yeah, I mean, today we have all of our product teams asking these questions like, hey, is AI the right tool to solve this particular problem? So we invested a lot of time in making sure every product team can run discovery around good potential AI use cases. And that is really important that we have product managers, designers, engineers. When they’re talking with customers, they’re always able to spot these good opportunities for usage of AI. And that’s what I think is the AI first approach. We’re not doing AI for the sake of AI, but this is an important tool in our tool set to solve a large amount of problems that are present in the recruiting space.
04:11
Matt Alder
And I suppose is the challenge there identifying kind of use cases that don’t exist yet. So ways that AI will kind of fundamentally improve things by solving problems that people don’t know they have.
04:24
Michal Nowak
Yes, that’s always the thing, right. I think this is the big role of engineers coming to the conversation with a customer, because engineers are living on the edge of what is possible right now. So I think the contribution of an engineer to that whole process is showing and showcasing what is just now possible, because oftentimes customer have a problem but is not aware of what technology can deliver and what technology can do to solve that problem. So I think this is the big role of all the engineers out there that should be joining customer calls, should be joining discovery conversations, and should support that process.
05:06
Matt Alder
Yeah, I think that’s a stunning point. And that’s. That’s exactly why I wanted to kind of really talk to you, because there’s just so much conversation about AI and so much speculation and so much stuff that isn’t really true, but also so much going on that people don’t know about. So, yeah, I mean, I think that kind of makes perfect sense to have people in the conversation who actually genuinely understand the technology. Because they’re building it.
05:27
Michal Nowak
Yeah, exactly. And they’re also like, there are some qualities of a problem that when you look at it, you think this might be a good use case for AI. So if you see something that is repetitive and done in High volume. Okay, that might be something good. When there is a problem, when you need to analyze a lot of different inputs, like availability of 10 different people to find a good spot for an interview. Right. And if you have like, if you’re thinking of generative AI, if your problem is like tolerant for imperfection. So there is, you know, you have a Genai LLM model, you can ask question six times and for the first five time you’re going to get the same answer and the sixth time you’re going to get the different one. These models are not deterministic.
06:17
Michal Nowak
And so if your problem is like tolerant to this, then you might have actually a really good use case for AI. So this is like one of the lenses that we use to identify these problems.
06:29
Matt Alder
It’s interesting because as you say, we’ve sort of been talking about AI for a very long time and gen AI has really accelerated that conversation. And one of the things that’s always kind of come across historically is whenever we talk about algorithms or AI or robots or computers, there’s always a kind of a sense that, well, hang on a minute, recruiting is this very human activity and replacing it with machines is not a good idea. But I think we’re seeing that AI is actually helping to build better human connections, isn’t there?
07:02
Michal Nowak
Oh yeah, it’s such a interesting dynamic right now happening. So you have one side of this recruiting space where we see AI agents building job descriptions and crafting job descriptions for recruiters. We have yet another AI agents building CVs for candidates based on these job descriptions. And we have like yet another AI agents matching these CVs or like ranking these CVs versus job descriptions. So like it is that weird dynamic where we as human beings are watching these agents have conversation with each other. And this is really like, I’m not sure where this is heading. We see like more and more companies struggling with fake candidates even on the interview stages. Like I just recently read about the company, like dealing with deep fake candidates on the video interviews.
07:58
Michal Nowak
So we have one side here where this is not heading in the right direction, I think. But then on the other hand, you have so many great use cases for AI that actually to your point, help to build that connection, Right? So if you think of scheduling an interview at scale, right, how many candidates never got an opportunity to talk with a company because like simply like recruiters couldn’t schedule the interview fast enough and now this can be greatly improved or there’s just like how many candidates never got any feedback from a interview process? Because simply like, you know, people didn’t have enough time to process that volume of candidates. Now we can use AI to build thoughtful messages and actually pass quality feedback to candidates, identifying skills within.
08:52
Michal Nowak
There’s a large talent pool and now AI can actually surface the great qualities, the skills of the people in that pool and make it more visible. So all of that, I think is that positive aspect of AI where we really build connection, shorten time to connection and make these connections more thoughtful. So yeah, I mean, and I’m excited about like developments in here.
09:18
Matt Alder
This to me is the most brilliant bit about this sort of current technology revolution. And also, you know, you make a really good point there because there’s all this conversation about the human touch in recruitment and all that kind of stuff. And from a talent acquisition perspective that makes perfect sense. But from a candidate perspective, particularly in some of the job markets we’re seeing at the moment, people aren’t getting any kind of human touch or any kind of communication. And in some ways it’s almost dehumanizing. So I think that technology is doing potentially a fabulous job here of making it better for everyone.
09:49
Michal Nowak
Exactly.
09:54
Matt Alder
The other really interesting thing that is starting to sort of really come through in recruiting technology that’s sort of being built for AI is that the ui, the kind of, the direction that the UI is traveling in is very chat based. So literally just, you know, talking to software rather than kind of pressing buttons and all that sort of stuff. For many people that will be a real sort of departure from how they’re used to using technology at work. I mean, how are you convincing TA teams that, you know, running an ATS or a chat based interface is a good thing to do? I mean, what are the advantages of doing it? What’s the kind of reaction to it?
10:32
Michal Nowak
Oh, this is such a big topic, an interesting one. It’s a, it’s a transition happening right now and I think it’s not going to happen overnight. But there are like lots of elements that will help in that transition or help identify the good use cases. So I think one important bit is showcasing the agent or the conversation interface in the relevant places in the ui. So we put a lot of focus in to make sure that the agent will pop up or show in the most relevant situation. So you have that AI agent that is just listening to the inputs and events in the system and then when it’s relevant we show them like so, hey, a candidate just declined an interview.
11:20
Michal Nowak
Would you like to, would you like me to ask for A different spot or hey, I noticed that your job is not getting any candidates. Maybe you’d consider advertising it in more places or changing something about the process. Or hey, we noticed a large drop off of candidates in a certain stage. So an agent can notice and spot these places and showcase and initiate the conversation. So that entry point is important. I think the other thing is how you present the conversation. Some of the recruiting flows are very complex and I think everyone right now try to use a mixture of a conversation and the dynamic UI elements.
12:07
Michal Nowak
So a chatbot, when asked about location, renders a map and you can click on the location or if there’s a simple form to be shown, like the tools right now can dynamically render the form and you can just type it in. Or if there’s like two options to choose from, they’re rendered as buttons, right? You don’t need to type it. So you need to recognize these situations and help to basically drive the conversation through that adaptive ui. It also helps agent to be less error prone, right? Because if you click on a button versus type or if you just pick the availability versus describe it is just much more robust and it’s in times faster and it’s a really hard work on the design end.
12:52
Michal Nowak
It’s just like a different breed of products and different way of interacting with an end user that you need to design right now. So that’s like, we’re all like spending a lot of time learning about that. So this is important. I think the other thing here is like the agents, especially recruiting agents, we want to meet users when they normally live, not necessarily inside the recruiting platform. So we want to make sure our agents live in teams or in slack or in WhatsApp. And so we meet people when they normally use a conversation and have that recruiting conversation also in the experiences or in the platforms that they’re by design, by default conversational based. So these are some of the things that we’re thinking about and there are also some use cases that are just so much easier when you have a conversation.
13:53
Michal Nowak
So imagine you have your entire week of interviews scheduled and there’s Monday morning and now you’re sick and it’s so much easier to type, hey, please move all my interviews to next week versus go there and just click through that. Laborious UIs to do that, right? So I think these are some of the areas we need to think about.
14:18
Matt Alder
Yeah, I think it’s a really interesting area and I’ve noticed that from having a 10 year old child the way that they interface with anything electronic is conversational straight away, normally voice based. I didn’t even know you could talk to the television remote control in our house and find what you were looking for. But I kind of learned that from my kid. I’m there sort of pressing buttons and scrolling through menus and he’s just asking it to put out whatever he wants. So it’s really interesting to kind of see that sort of development and where sort of user interfaces are going to evolve eventually end up.
14:51
Michal Nowak
Yeah, I got the same with my kid with like conversational conversation with ChatGPT. My son is like 7 and he’s, hey, figure out the trivia of my favorite like, you know, TV series and ask me 10 questions, right? Or like tell me a story combining Harry Potter and Lego and something else and like, you know, so the possibilities are endless and kids are really like doubling down on it. It’s really interesting to see them interacting with the technology.
15:21
Matt Alder
It must be amazing to be that age and dealing with this kind of technology because it just seems normal sometimes to us it seems a bit like magic, so. Absolutely. One of the other things I want to talk about is AI, obviously, as we say, it’s magic. It’s amazing. But it also does raise some kind of fundamental problems as well, particularly around ethics. And there’s obviously lots of regulation coming down the pipe, particularly in Europe when it comes to applying AI to making decision in recruiting these kind of things. How are you addressing all of that?
15:53
Michal Nowak
So this is, we feel like a weight on our shoulders when designing it because like, if you think about it, there are like AI systems right now that score candidates. They will attach a number to a candidate of like how well that CV fits a job description, right? Or we going to be coloring parts of the CV in different colors. Like we’re going to be identifying relevant skills or missing skills within the cv. Right? So even if there is no automated decision making and we’re never deciding for recruiters, we’re going to be influencing the decisions, right? So these AI systems only by showcasing some elements of a CV or like, you know, giving a higher score here and in the other place, they will influence the decisions.
16:46
Michal Nowak
So it is extremely important to make sure these systems can be trusted because we’re touching into fundamental rights of equal opportunity. If you think of how these systems are deployed, it’s just happening at scale, it’s happening across millions of candidates. This is extremely important. And there are a couple of things here that I think needs to happen. Right. So first of all we always try to design these systems together with customers. We have amazing customers. They’re really passionate about the tech and they just spent countless hours with us, giving us feedback, giving us guidance, and telling us, hey, what are they comfortable using and what do they need to see in the system just to make sure this is safe for them to use?
17:43
Michal Nowak
So I think this is really important to always design this system together with the user, together with customer, and make sure we build something that they’re comfortable using. And that’s just by principle. One of the most important thing then the other thing is explainability of the AI system. It cannot be magical. We need to always try to showcase how the system made those decisions and how we arrive to these conclusions. And this is what we do for our own system. So also whenever we work with partners, we demand explainability for these systems just to be able to showcase all of that. And then you need to test it. You really need to put a lot of thought into and effort into testing these AI systems. And yeah, if you think about it, there’s just a lot to test, right?
18:39
Michal Nowak
Like if you want to test for bias right now, we need to be like, there’s New York City Law 144. We need to be tested against gender and race bias. But there are so many other aspects of bias, right? Like it can be age, disability, institutional bias. Right. So you need to be aware of that. And. But there are so many other aspects, right? Like agents can be toxic, they can use profanity, like how can we make sure this isn’t happening? Or they can use stereotypical. Stereotypical behavior or unethical behavior. Right. So there are like all these different aspects of agent trust. And like we’re working really hard to build test suits for that just to make sure we verify this before deploying the systems. Yeah, so there are some of the things here, but yeah, I mean, I agree with you.
19:32
Michal Nowak
This is extremely important and something that we need to really put a lot of focus and thought into.
19:39
Matt Alder
So a final question for you.
19:40
Matt Alder
What does the future look like?
19:42
Matt Alder
What’s the kind of vision that you’re kind of building towards what might be possible with this technology in a few years time?
19:50
Michal Nowak
It’s a hard question, right? I mean, no one really knows, but I’m pretty optimistic about how the future will look. Like I. Recruiters are so understaffed right now, smashed with admin, like overworked often. Like this is what we hear from most of our customers, that it’s really tough out there. So I, I would imagine recruiters of the future will have some breathing room and they will be able to think about their work in more creative, more strategic way. They will spend a bigger portion of their time thinking of what are the strategies to attract people, how can we leverage the existing talent or design internal mobility programs, how can we design better evaluation process for our candidates? And I think I would love to hope that’s going to be their world and that’s going to be the shift for them.
20:55
Michal Nowak
And I also think candidate world will be also improved by that. Right. Instead of just engaging in that tedious process for the admin around applying, they will spend more time actually having a real conversation and spending more time showcasing these their skills and basically they will better matched to opportunities. Right. So I would imagine this world will become easier over time. So that’s my optimistic prediction of how it will look like.
21:30
Matt Alder
Michal, thank you so much for talking to me.
21:33
Michal Nowak
Thank you so much for having me. It’s been awesome.
21:36
Matt Alder
My thanks to Michal. You can follow this podcast on Apple Podcasts on Spotify or wherever you get your podcasts. You can search all the past episodes@recruitingfuture.com on that site. You can also subscribe to our weekly newsletter, Recruiting Future Feast, and get the inside track on everything that’s coming up on the show. Thanks very much for listening.
21:59
Matt Alder
I’ll be back next time and I hope you’ll join me.
