LaunchPod: Mykola Konrad
===

Jeff: [00:00:00] All right, Myk, welcome to the show, man. Good to have you on

Myk: Same here, man. I, I love being here, so thank you very much for having me

Jeff: Yeah, yeah. It's, uh, it'll be a fun day. I'm excited to have you on. , you're CPO over at Euna Solutions right now. You've had a, you know, long, , career where you've kind of seen a lot of things, and I think one of the big things we're all running into right now is just this, massive rate of change.

Obviously we all, uh, know those two little letters of AI, but, , it is leading to change in several different places, not just in coding, but, you know, UX teams, what does UX look like? And even just like how can customers absorb all the work we're doing? And I'm excited to talk about that from a kind of standpoint of maybe a little bit more regulated industries , and kind of the areas that Euna Solutions is selling into.

, But maybe before we jump into that, just a quick TLDR , on how did you get to, you know, being chief product officer over at Euna and, what did that look like?

Myk: , Happy to provide that. , So started off as an electrical engineer, then I became a product leader. I've been doing this for over 25 years, and my background are places like [00:01:00] Microsoft, Avaya, Ribbon, Brightcove. , And I've sort of pivoted from hardware to software and for the last, let's call it over 10 years, B2B SaaS.

And so that's where I've been doing product management and product-type activities

Jeff: Nice. if I got this right, over at Euna Solutions, you don't just have the product team, you also are responsible for UX team, correct?

Myk: Yeah, correct. And the way that we are organized, I know in some companies the UX team is under the development organization. In our case, we do have a separate CTO. , But in Euna, that's not the way we're organized, and UX reports into the product, you know, chief product officer and is part of the three-legged stool, but underneath product

Jeff: Mm-hmm. So this is the thing I, I, I really wanna kinda hear your, end on or your thoughts on is, you know, I've, I've spent the past year traveling across the country and talking to, product leaders and senior product folks from San Francisco to Salt Lake City to Raleigh, North Carolina, so kind of, you know, [00:02:00] across the board.

, And no one can agree on what's going on with well, products in general, , but especially, you know, UX and, is UX headed for a tough road because, you know, if it's all agents, who cares? Is it gonna be AI designing all the UX going forward? Or, or do you think there's still , kind of some human element that's going to be important here where the robots won't ever fully be able to replace, you know, the, the taste that a, a well-learned, uh, human can bring to that discipline?

Myk: Yeah, it's a great question or a great, I think, topic to discuss. And, uh, actually I was at the AI Boston Week, uh, last week, and I actually asked a few UX designers that happened to be attending some of the different sessions what their thoughts on this were, . So this is a, you know, not just a Mykola Konrad opinion, but also trying to gather because I'm learning, I think we're all learning how to, how to navigate through this.

So I believe, and the way that we are looking at this or I'm [00:03:00] looking at this for my team, is that there's gonna be certain aspects of UX/UI that are protected and when I use the word protected, that are value-based activities that - need to continue to happen over the next couple years at least.

Uh, you know, I, I'm not gonna prognosticate more than three years from now.

Jeff: That's even a long time at the speed

Myk: It does, yeah. It does seem like a very long time, you know. But, uh, listen, you know, ask anybody about a roadmap and it's gotta be at least three years, right? So everybody's gotta know where they're going, even if they don't really.

, But there are other activities that right now -- we are actively moving away from a dedicated UX/UI person and trying to move, somehow meld them into either an existing product owner, product manager, or even development - owner, so that there isn't that need for that activity happening in the UX team.

And so to give you some examples, I believe that activities such as, , design system upkeep, right? So keep... Let's get the design system, [00:04:00] let's get the new component in there. , First drafts of a UX, some of the early exploration, the prototypes, getting from a PRD to a prototype,

The activity still happens, but it will move to a different persona within the organization. And right now we are actually having some product managers do that, and we're also having some developers do that 'cause I'm not sure exactly 100% where it's gonna land. And so right now we're following this mantra of, "If you can, then do."

So what that means is if you as an individual believe you have the aptitude or the skill set or just the desire to try to, as an example, prototype something using, you know, we use Anthropic, so Claude Code or Claude Design, then do it, right? Don't wait for the UI/UX team to help you out, just try it. , So we're on this, "If you can, then do" mantra right now.

I think we're gonna eventually move from there, , to a, "Well, it's your job. Do." So those are some examples of things I think are definitely going to move. things that I think stay, at [00:05:00] least for the, let's call it next couple years, are, direct customer contact. So in a... In-- You, you still wanna do an interview.

I've actually had the experience of being interviewed by an AI agent. It's okay. There are just questions that I was, I was thinking, "You know, you should be asking me this instead of that."

And if I were, for example, speaking with Jeff, I'd say, "Hey, Jeff, you know, let me go ahead and answer this," but it's too scripted that it's hard to do that. so I believe that the customer contact, , the moderated sessions, the final call, right? So the, "Hey, this type of object or this type of flow is going to work best for the user," , that I believe is still gonna be in UX for now.

in case of what we do, which is we have some high-stakes flows, things like compliance flows or audit trails or payments, right? We actually have a payment product that we own the rails and the payments are, crossing through us. Some of those [00:06:00] aspects, you're gonna want somebody that's really deep into that UX to make the final call.

Yeah, Claude Design can give you a first pass, but you're gonna want somebody else to look at it. , Accessibility, right? Things that are more regulated, I'd want the human there for now. , But my suspicion at this point is that, hey, there's certain things that are obvious. Like I was just saying, the design system, the first drafts of the UX, that's definitely going away, and in, in our case, going away literally, over weeks.

Like, a month ago, somebody was still prototyping this. Now we have an agent doing that. and then as AI gets better over the coming years Some of these issues - will move into the, "Ah, we don't really need a live UX person to do that." I can believe some of the interviews will eventually go to that, , to that area.

I can believe some of the insights, , the final insights will go move , to that area. But the accessibility, that's still gonna need some human [00:07:00] being in the loop. Some of the regulatory, , compliance areas, still gonna need a human in the loop. , So, the bad news, and this is, again, this is now my opinion, is if you're a starting off graphics designer and you wanna get into UI, I think it's gonna get increasingly more difficult to do so.

the good news, if you're somebody with 15 years of experience and you sort of have an understanding, as an example in my space, which is the B2B SaaS gov tech space, of how a, potential employee that is doing grants management or budget management would navigate through this because you've been doing this and have talked to people for the last 20 years or 15 years, there's real value in that, and there will continue to be value in that.

Jeff: I think there's two pieces in there I kind of want to dive into. One thing is it sounds like, you don't necessarily think UX as a role is going away, which I, I tend to agree with. I think most things are not maybe going to be fully taken over by AI, at least in the, you [00:08:00] know, we'll call it , the somewhat near timeframe, whether that's three years or, 18 months or five years or, whatever.

, But , the system in which these things operate, as you're seeing it, are probably not just going to shift but are shifting where, before you had to have kind of production line UX and, and designer people who could do every last knit on any prototype or anything going into production.

And kind of the same way that ClaudeCode or Codex or any of these tools bring the bar down on who can do more, right? Like if you look at two years ago, I couldn't write more than a couple lines of code. I could edit , a little bit of code if I kind of knew what it did, but that was kind of the, the top of my limit.

And now internally, I've shipped several applications we use, , internally across some teams at LogRocket, but there are kind of, you know, certain limitations, some security review and stuff I still want someone to look at, but I can do all of that. Same way you can, you can build prototypes, you can build to a certain level, it sounds like within the design and UX kind [00:09:00] of parameters and, and rule sets , and setups that you guys have configured over there.

But you are still gonna need that kind of human in the loop review where it gets, incredibly important to be accurate or high stakes , is that kind of how you're looking

Myk: Yeah. Yeah, I think the role is there and I-- But-- So let me give a couple points there. One, you know, we have some contracts with some big states, , and provinces in North America, and for the work that they want and need, we need a human being to interact with them and say, "Okay, what is the flow that this grant administrator wants?"

They will not accept, an agent - interfacing with them. I wouldn't even call it regulatory, it's partly the way that that particular government entity needs and wants to operate. They will not accept an AI doing it for them. They want a human being doing it. , The other aspect of that is that given, , UX slash right now UI, , person, I think will be either be [00:10:00] able to do more, meaning, "Hey, I could have pumped out, an entire, you know, 40 screens worth of work, , prior to AI.

With AI, I can do 100 screens worth of AI work, , and I know it's gonna be good because it meets the criteria that I've researched using AI, and it's gonna hit, it's gonna hit this persona and give them the flow of the data that they need." so there's this aspect of same person just doing more. The other aspect of this is, hey, you know what?

The product owner, which before would never touch any coding , up to there, will actually go ahead and code the first few screens and show the developer it's got to look something like that, and then the frontend developer takes it from there. So I can see a world where there's more work because there's more output and more capacity, so there's fewer UX/UI roles, but they're just higher capacity because they're, they're just pumping it out.

But I can also see a world where some of that work is just done by other functions. Right now, , where I believe what's [00:11:00] happening is that more of the work is moving to other functions. Therefore, I don't know, just to use round numbers, if you had 10 UI/UX developers, maybe now you need seven, because some of the lower level work that doesn't require quite that taste and that eye is now going to the designer or to the product owner or what we now call product owners.

There's a separate discussion what happens to that role in this future world. , But I do think it's gonna, it's gonna move. And when I think about future, future, which could be to your point two years from now, I do believe that we will not need the triad fully. We will need two in a pod and then...

Sorry, two Ps in a pod, right? And then maybe you'll need an augmentation, somebody that's just checking in to make sure

Jeff: I get like right now, and this is not just you, this is, you know, every company on earth that is not a brand new startup, you know, every established company there's a team. And then if you look down the road, whatever, you know, meaningful amount of time, it's probably not that long, [00:12:00] like you said, there's a future state of that team it looks different.

How are you, , getting, you know, that current team ready for the future? What does that look like?

Myk: Yeah. So, , in the throes of that now, and it's not an easy answer. So first off, change management 101, right? People don't like change. - Most people don't like

change. , , I saw a paradigm, a framework for this once a while ago that said, "Listen, nobody likes change." It just some people get through the don't like it phase quicker than others.

And you have some people that'll stay in the don't like it phase and the resistance, to change for, you know, I'm gonna use a number six months and others because of their nature, just inherently how they think about things, they'll get through it in a week. So the goal in change management is not to say to anybody, "Well, I don't know.

Why do you hate change?" No, like embrace change. , It's not even that. It's let's get through this. We're all gonna hate it. Let's get through this really as fast as possible, so we're on the other side of the change. so with that, you need the clear goal. So in our [00:13:00] case, well, what are we using?

, What's the tool system that we're gonna be using? , We're Figma Make Shop. , We do use Claude Design for some of the prototyping. So we're sort of this two tool system right now. great. So let's all make sure everybody understands that and how to use that. So there's some training involved.

Then what is the new way that we are going to approach handoff from our design agents to, product and to our software design factory? let's Visio diagram. Yeah, we still used Visio to, to diagram that out, , so that everybody understands here are the steps, here's what we expect. And then a conversation which right now is very loosey-goosey on, well, how long should this take?

designing something like this used to take three weeks. I don't know. Maybe we should say it takes three days now. Let's try that and see what happens. , And what I've found is that there were some people on our current design team [00:14:00] that just jumped in ,

I'm gonna use the word monitor this, but I, I wasn't necessarily looking to monitor this, but you can get an insight into it by looking at token spent, , and you know, how much, how much activity is there. , In others you could tell to they just weren't into it. They weren't moving through that change fast enough.

, And if you have the-- I'd say as a company, if you have, , the ability to wait for a while, everybody will eventually make it through this window of change management. For some people it might be a year. If you have the luxury of waiting a year to get people through this change management and with the training, you know, great, like you can wait a year.

If you only have two months or three months and you need to see people popping out the other side of this, then you need to really understand, well Is this person, he's not gonna make it in my timeframe, or she won't make it in my timeframe. Maybe we need to move on. , It's a tough decision.

It's not a fun exercise

Jeff: it's funny you, you bring up kind of that you can, , look at token spend and, and [00:15:00] figure that out because, , , a lot of companies famously were just pushing token spend as a measure of productivity.

they all kind of realized like, "Oh shoot, I just blew my entire budget in four months," , and you know, that whole fun little rigmarole that, that a lot of companies went through. , But I do think there is a point there. , Token spend unto itself , is not going to measure if someone's getting it or not, or if someone's moving or not, but it is a, a piece of a bigger picture in an important way.

If someone is just, you can tell using nothing or, you know, a couple thousand tokens a week, - they're probably not changing. If they're using a ton, , that's a good signal that they're trying. 

Myk: that's a good--

Jeff: deeper,

Myk: Correct. There's a signal and there's-- Yeah. And then the other, the other thing that, you know, we've done or, or I've done is actually as other parts of the organization, and I'm talking product managers , or developers, because , in our product management organization, we're becoming agentic in terms of how the epics are written, [00:16:00] how the stories , are then derived from those epics and pointed out.

And there's certain s- again, since we're using Claude, there's certain skills we develop for how to write a PRD. Some of those basic things that still are happening. And you're giving that to the development team, and the development team has their own software factory, and it's the agents are taking and developing, and QA is going back and testing.

So all this is becoming agentic. Well, you ask that team, "Well, I don't know, what's your biggest, , sort of point where things are stopping right now?" Right? Where in this smaller team of six people that are working together every day, where's the pain point? And, you know, I recently had this discussion where, well, you know, it's probably the UX and how quickly we're getting the, uh, new designs.

, So then you go back. Well, , what's taking so long there? Is it because the user research isn't done yet? That would be a very plausible explanation because user research requires users, which requires customers to actually respond to you and talk to you and get on a podcast like this and actually have a discussion.

, [00:17:00] Or is it, no, all that information is there, so it's not a data in problem, it's how quickly the data's getting out problem. , Right? And so once the teams start getting faster, what you'll find is , just go talk to them and say, "Listen, you guys know sort of what you're doing," or the team is, you know, do a 360 review and , that points to where the gaps are right now.

Is it a process gap? Is it a data gap? Like, what is keeping you from going faster? And specifically with UX, what is it that's keeping that from moving faster?

Jeff: I'm curious though, like, 'cause you brought this up, , on the engineering side, you know, they might have their own kind of software delivery processes or, or software factory.

What does that look like internally now? Or how are you kind of coordinating across those teams to ensure, as UX and product and everything on, on your side of the house are, speeding up or trying to speed up, that that's not throwing into disarray what's going on over there? Or vice versa, if they're speeding up faster, how does that handup work to, keep the gears turning kind of somewhat in sync?

Myk: Yeah. So first, let me preface this with saying we're still working [00:18:00] through that.

Jeff: Yeah. I think everyone is, if it makes you feel better.

Myk: But what, what we are, , attempting right now, and I think we've got two separate flows. One I'm gonna call AI-enhanced, and the other one we're-- that we're trying, which is full AI. AI-enhanced is where we have agents on, let's call it my side of the fence, the product side of the fence, that are writing the epics, decomposing the stories, and pushing them into, in our case, Jira.

So you end up , with an epic and a lot of stories underneath that. Some of those stories might actually contain links to prototypes or some, some of the epics, sorry, might actually contain links to prototypes that the, uh, PM and PO have done. The dev team looks at that. The dev manager will look at that, or the team leads will look at that and say, "Okay, I, I sort of understand what's happening here."

, And then push it from there to... Or if there's any questions, maybe go back to the PM, , but then push it into GitHub and into their agent factory, which will , [00:19:00] start the work. , The agents will pick up, you know, there's multiple agents. They're all orchestrated by an overarching architecture agent that will go start picking up the stories one by one by one and start developing code, which then goes to QA.

And in the meantime, QA has also looked at the stories and developed, , tests, , for it, and that can be... Right now it's, I think, probably more human than agentic, but the goal is for that to be developed by agents, and agents test a-against the stories and then go back to the PM and say, "Great, here's the checklist.

We did all this." Uh, or sorry, the PO, "We did all this. Do you wanna actually check the final output? Like go take a look at what it looks like, , you know, run through it yourself, , to give the final okay." So it gets then pushed to the pipeline and then out. That is what we are doing right now with a few of our products.

Where we want to get to, which we're testing, is - Epic still happens, stories still get pointed, prototypes might still get done, but then Jiras are still there, so we can [00:20:00] go back and look at what happened. But then the agent actually opens up, so a product agent opens up, , GitHub and pushes some information, actually might even be code, into GitHub, into the repository for that specific product context.

And then the dev agent automatically goes in there, grabs that, as well as, as context from Jira, and starts developing from there. , And so there's almost no human in that loop. , That is a little scarier, , a little faster, and we aren't there yet, but I believe that that's where we're gonna end up. And still at the end of the day, the QA does QA, the PO will still check the box and say like, "Yes, this is what I thought it should be."

, And then if it's multiple epics, the PM will come in also and check it to make sure it's okay. So I think even in that kind of situation where it's fully automated, almost autonomous, there's still going to be a check, dev will check the code probably, is what we're thinking right now, [00:21:00] but the actual outcomes will still be checked by product.

Jeff: Yeah. I haven't seen a lot of companies yet where they're, they're fully kinda just YOLOing it into production. I feel like you see a lot more where they're-- the, the automated kind of, you know, maybe issue fixing is, is starting to pop up and you're getting companies, , doing, you know, , pull requests and, and stuff like that automatically before a human steps in.

I don't think I've seen a lot of companies , at large scale at least committing to production without a human, , looking at it. And the kind of point there is to give the humans time to spend more time on the big things that are gonna be, , more either forward-thinking or more complex to think about how you solve those problems.

It's like the small little ticky-tack stuff, , that is getting less and less attention, so you can do more of it. , , It's the stuff that humans didn't like doing and didn't really value add before, it's just you had to rote do it.

Myk: I completely agree. And I have this thought, which we're currently thinking about within the halls of Euna, that eventually we will get to a way for the product team, maybe with [00:22:00] dev, uh, dev help to list the epic and say, "Listen, I think this epic is 90% or 100% agentable, so doesn't need any human in the loop.

This one is 50/50, so we do need some check at the end. This one's only 10% agentable for whatever reason, regulatory, , depth of nuance of the code. That one 100% needs a human in most steps." And so, so you have these different tracks

Jeff: And my, my question on all those kind of delineations , , kinda comes back to over time,, how much of that moves up and up and up in percentage-wise that you can just toss to the agent? And where do we asymptote and, and do we? , Which is kind of a scary question. , , but, , you know, speaking of prioritization, I think we have to give you up in a minute here. , so maybe this is a question we leave for, for next time we can get together and talk or, or maybe for people to kind of ruminate on, , after the episode.

But if you are moving faster and faster and faster, and that in theory means you're probably shipping more and [00:23:00] more and more, , or at least across the board companies as, as a average are, customers are still human. this is still humans using a lot of these outputs, and they can only kind of ingest and, understand so much new product.

And, and, you know, Euna is not operating in a bubble here, right? LogRock is not operating in a bubble. Anthropic puts out 18 new features, I think, since we've been on this recording today. It starts to become how do you work with your customers to make sure that they're, understanding what's coming out and that they're understanding and able to use it, and that it's not just usable, but they're having success with it.

And that becomes a whole other - problem that, that maybe we can tackle a different time.

Myk: Yeah, I'd love to have that, that discussion. I actually think the asymptote to, to what you said before is going to be the rate of change that a customer can actually accept. And in different industries and even different customer subsets within an industry, I think it's going to be different.

Like we are seeing that right now. Like some of my customers, they are in a one-year cycle. Even if I had the, the best, most value-added features [00:24:00] in the world that would make their lives so much simpler, it's a year from now, right? That's what it is

Jeff: Myk, it, it's been a pleasure, man. I appreciate you coming on, , going through all this and, and kinda going into how the team over at Euna Solutions is thinking about this change in UX and, and UI design and, what that looks like over time and where, you know, how do you shift a team?

How do you get them ready? , And what do we think the future looks like here? , Hope you had a good time. It was a blast having you on. If people want to ask more questions or, or dig into you more or just wanna, you know, say hey, is LinkedIn a good place to reach out?

Myk: It is. , Love to have the discussion because the more we talk, the smarter we all get. And thanks a lot for having me. Really appreciate it

Jeff: definitely. All right., Myk, Mykola Konrad on LinkedIn. , It's been a pleasure, man. , Talk soon. Have a good rest of your day. Thank you