DPM – How Calendly Is Using AI to Deliver Without PMs

[00:00:00]

Galen Low: Hey folks. Welcome to our session on how AI is helping remote First Scrum teams deliver complex software releases without a delivery manager. We do events like this once every few weeks.

It used to be monthly. We’re doing them more often now. Uh, and it’s really our way of having our members and our VIP guests engage directly with some of the experts who contribute to and collaborate with us here at the Digital Project Manager. Uh, for those who don’t know me, I’m Galen. I’m the co-founder of the Digital Project Manager, and I will be your host for today.

For better or for worse, uh, I’ve also got with me the CO CEO of Rails Wear. Sergei, Olav, and Rails Wear’s product directors. Hey guys, Julia Kova. We’re gonna do formal introductions shortly, but I thought maybe we could actually lean into a little tradition we have here. Um, maybe you could just let us know in the chat where you’re joining from and maybe just like what your biggest work challenge has been lately.

Uh, it doesn’t have to be a long description. Even just give me like two words that sum it up. Maybe like note taker hallucinations, back to back releases, or doom scrolling about how AI agents, uh, that weren’t supposed to take your job actually seemed to be taking 50% of all entry level white collar jobs and maybe even replacing schools.

Okay, maybe that was more than two words. Uh, while you do that, I’m going to go through a little bit of housekeeping for today’s session. Uh, so I should let you know that today’s session is being recorded and will be available for members of our project management community shortly thereafter. We may also use some clips for our website and on our social channels.

Um, but just so you know, your cameras and microphones are off by default, so you will not appear in the recording. Um, and also we don’t think it’s rude to keep a conversation going in the chat. We think that’s part of the experience. The chat will also not be captured in the recording. Um, and we are not the only experts in the room.

So if you’re sharing ways that you’re using AI to help your team get more done, develop great products. I wanna hear about it in the chat, uh, and we’ll engage over there as well. Uh, and also I should say that we will make some time at the end for questions from our DPM career builders and members of the audience.

Um, just post a question if you’re, if you’re a member, post a question in the live events, uh, channel in our Slack space. Uh, also feel free to, to throw them into the webinar chat here, um, and or the q and a section. Uh, and we will get to as many as we can at the end. Um, also, we’ve got a few VIP guests in the audience today, so if that’s you, welcome.

This is just one of a series of monthly sessions we hold for members who also get access to a number of other benefits, including our entire back catalog of session recordings, our library of templates, resources, and mini courses, as well as our flagship certification course, mastering Digital Project Management.

You can join the fund by heading over to the digital project manager.com/membership. Woo. All right. Without all that information, we can, yeah, I know. I try and go through it as fast as I can, so I don’t spend the whole time talking, uh, blah, blah, blah. But, you know, it’s important to set the context. I guess I’m seeing a lot of, uh, really cool, uh, things in the chat.

I. Some people from really interesting places, uh, and also some challenges. Yeah, I, I, I hear that there’s a note here about a company expects me to use AI but doesn’t explain how, and none of our tools allow any integrations. I think that’s very topical. It’s one of the things that we want to cover today.

Um, so let’s, uh, let’s, let’s dive in. Today’s session is all about how remote first development teams are using AI. To tighten up the way they deliver impact through their digital products. Um, maybe we can just start by meeting our panelists. I’m gonna pick on Julia first. Uh, so we’ve got Julia Ryzhkova. Julia is the product director at Rails Wear.

She’s got 15 years of experience leading product, people on process across SaaS and services. She’s known for scaling [00:01:00] products from MVP to millions in a RR, uh, aligning customer value with company strategy and mentoring teams to deliver with clarity and purpose. And Julia, I wanted to get into this because, uh, you just wrapped up like this three day business simulation game at Thera House School of Business, where you’re doing your MBA and you ran full on businesses from scratch, including marketing plans, production flow, sales ops, finance reporting, leasing, procurement partnerships, and even mergers.

Uh, you were the CEO for your team. I’m just wondering if you could share, I don’t know, a surprising takeaway from those three days.

Julia Ryzhkova: Mm. I would say that everything is pretty much, uh, in simulation, uh, close to the real world. Like, uh, you need to negotiate with your, um, ex-friend. I would say during the simulation they were not, uh, not friends, uh, like in each million counts.

Uh, so like about discounts with, uh, tax offices to actually pay your taxes. Not the end of the year, but the beginning of the year. So everything is about, um, some creative way [00:02:00] of how things are doing. And we didn’t use AI because like everything was simulated as a board unfortunately. Um, however, that’s probably the, uh, most interesting thing.

Uh, I came up with,

Galen Low: I like what you said in the green room as well, where you’re like, this is all really interesting, but the way we work at Rails wear is so different that it almost doesn’t apply. We were blowing people’s minds.

Julia Ryzhkova: Yeah, it was in another, it was about time management because, uh, usual time management in corporate environment, uh, can be applied in, uh, else because of the maturity of the team.

I would say that we have, um, it’s, uh, we have a flat structure. I, I mentioned it during our previous podcast, but of, of course, there are many people, uh, who are new here. Uh, so we don’t have linear managers. Uh, I have 14 people in my team, which is mind blowing, like usual, you know, that, like Ger said, that it should be from five to seven and that’s it.

Uh, so I have 14 people in my team that I lead. Um, and still I manage to work, uh, 40 hours per week. [00:03:00] Uh, something like one hour for operations per day, then three hours for meetings and everything else is dedicated to strategy. Culture, those kind of things that, uh, usually according to theory, you can delegate and manager sexually should focus on, but we never have time to do that.

So that’s exactly how it works in. And, um, at the same time, during KB courses, of course they suggest some approaches and practices that are more common for huge corporate companies. Uh, like, um. Schedule seven minute, uh, meeting with your team for the next week so that they feel the pressure, uh, like how your time is precious.

And it’s totally like, uh, I can’t use it in the roster that was in, I definitely, so interesting and pretty much during calls, the courses, it was like that because I tried to. Understand like how to, uh, use this in the in in elsewhere. And, uh, their, uh, lecturers also [00:04:00] learned a lot from me. How else? Because they only heard about Spotify, for example, model about flat structure, but they didn’t, uh, talk to real people who actually use it in the real world.

Love it. So it was like beneficial collaboration, I would say.

Galen Low: I wanna get into all of that as we go ’cause it’s like super interesting. I think it’s germane to like why I wanted to talk to you guys. Uh, I thought maybe I’d, uh, also introduce Sergey. Uh, so Sergiy Korolov, he is the CO CEO at Rails Ware, the co-founder of Mail Trap, coupler and Titan Apps.

He is a product first tech entrepreneur. He spent almost two decades turning ideas into high growth SaaS tools. He’s passionate about scaling with clarity, empowering teams, and just building damn good products. So Sergei, happy to have you here. I, uh. I understand that you’re in the middle of just like a few pretty massive product launches, including a big upcoming release of automation features and mail trap, an email delivery platform.

Uh, what, what’s your best tip for handling the stress of orchestrating such high stack, high stakes, high complexity projects like this?

Sergiy Korolov: Well, uh, it works [00:05:00] to say that, uh, Realware is a product studio. It means that we manage a lot of products. In parallel, there is a consultancy wing, but there is also a joint venture wing and there is a own products wing.

And, uh, so in total, I don’t know, we have about 20 products running in parallel, which allow, it’s, it’s a complicated thing, but it gives us a possibility to experiment. In different projects, different contexts, uh, of different sizes and different kind of industries. Um, it gives insights in one product and then allows you to apply those insights into the other product.

How to orchestrate all that. It’s of course the very mature team, as Julia mentioned already. Uh, so you cannot run all such like such a big team without relying on people who lead certain directions. Julia, for instance, uh, she’s running a very, very important client for us. Uh, company Trade Zilla. She’s a product director there.

[00:06:00] And, uh, she’s managing, uh, the context is very, very well. Uh, and of course she’s contributing to heavily contributing to Product Management Guild. So in our company, we organize every, every work through the guilds and squads. So the guilds is, uh, professionals, right? So you have guild of product managers, designers, uh, analytics engineers, uh.

Then you have Quas or teams, uh, which are kind of, uh, vertical integration of different roles, such as like it’s, let’s say, MEL Trap or Capar products, uh, which cover all the needed roles. So it’s more than 40 people in every team, uh, which represent, uh, and there are representatives from different guilds.

And so the knowledge we get from one product, uh, distributes through the guild, uh, to the other product. So the some learnings that you do in MailTrap product management, you find some interesting insight and you share it with your colleagues, uh, uh, in product [00:07:00] management channels or some demo sessions.

And we spread, uh, this information and experiment, uh, with such, uh, and, and kind of share the insights of the other experiments in the other products. So that’s, um, so, you know, it’s like flood structure, but it’s not like a flood structure. It’s actually a graph, uh, kind of a lot of, uh, different nodes that are interact with each other.

With their own rules. Uh, and those rules are very flexible. If we find that, uh, some cooperation between, uh, certain team members is efficient when there are like team of five, it might not be already efficient when there is a team of eight, and so you need to rebuild the structure and adopt it, uh, to the new team size.

Um, yeah, so, but I, I can go really deeply, uh, in that, so you, you should stop. We’ll get into that as well. Yeah,

Galen Low: no, we’re good. We’re good. You should stop me at

Sergiy Korolov: some point.

Galen Low: This is the interesting thing, right? Like I love that language about like the squads and the guilds. You know, you have this sort of [00:08:00] knowledge and nimbleness to have 20 concurrent projects and like these teams that are actually quite large, just like moving, moving quickly, doing things concurrently.

Um, I want to get into it. Uh, but first maybe let me just tee this up, um, because I. I’ve made this kind of click Beatty, right? I’m like, oh, look at all these. Yeah, make it a project manager. But honestly, for me it’s about this, it’s about the fact that, you know what, like not every team effort that has a deadline has a project manager.

In fact, in the world of SaaS product development, uh, and scrum based delivery, it’s actually probably more common for there to not be a project manager. So sorry to my audience of project managers, uh, but meanwhile. AI technology has been allowing development teams to do even more, uh, by helping them stay organized by streamlining communications, and quite frankly, by reducing the need for things like meetings, uh, and some of the overhead of, uh, of, of project management.

Um, so if some of this overhead is getting supported by ai, is maybe the globally distributed self-managing team about [00:09:00] to get like a second heyday as the pinnacle of digital collaboration? Or is it that the bar is already being raised for the level of complexity that might require like additional delivery leadership support, or is it maybe both?

So that’s kind of what I wanted to explore today. Um, like Julia, I had you on the podcast like it’s got, it was probably a few years back. Um, and I was just floored by the way your team works. Um, because it is flat structure, it’s nimble, um, and it’s very self-managing. Not to mention it’s distributed across time zones.

Um, so yes, I did make this title kind of click Beatty, um, but. At Rails wear, like this team, the team behind Calendly, the team behind Coupler io, the team behind Mail Trap and and Titan Apps, um, you guys have always been a remote first multi-time zone team delivering software without a project manager.

And I just thought I’d start out by asking like, what is it about your team culture that makes this like distributed configuration a benefit, uh, rather than a disadvantage? Like what are some of the things that your [00:10:00] team does that makes the work truly self-managing and truly remote first?

Julia Ryzhkova: Um, before I answer, um, as you mentioned, I’m currently working on my MB dissertation and I did the research, so I was also wondering like what is special and also that actually make it work.

Um, and uh, actually it happens that we’re not unique and the dissertation is dedicated to how, uh, in hybrid, remote and onsite environments, uh, uh, companies are doing it. Companies are working on with innovations. We didn’t find, by the way, any like, big, fully onsite company right now. It was one of the, um, things that we discovered during our dissertation because we simply couldn’t reach anyone who works like fully full on site unless this is a real small, uh, local IT company.

And we were interested in, uh, mid-size and, uh, large corporates. So currently all. Everyone are either fully remote or hybrid, uh, setup. Uh, and [00:11:00] for fully remote teams, uh, things that are common and also we use it in Railsware is exactly what I mentioned, the maturity of the team. And it starts from the culture in the core team that were at the beginning, uh, uh, as a foundation of the future company.

And then, uh, hiring processes and the working processes are shaped in a way to keep this, to share this culture. For example, I asked people during, uh, interviews, uh, okay, but how, how do people know that they can propose some changes? How do they know that? Uh, they can raise question, like challenge the status quo And they say they just do because it’s a part of the culture.

We don’t tell it, we don’t have policy. You can raise questions. It’s just that we hire mature team and during convertings, they, they see that that’s actually how it works everywhere. And then they know that they can actually do it. Some of the things, uh, like specifically we were. Um, I know that our audiences also is interested in specific things and tools and approaches that they can [00:12:00] use.

And, uh, that was one of the part of the questionnaire I also did during dissertation. Okay. But what exactly do you do? So, uh, some of the things are common. Um, we use Slack, uh, and we have public channels, and it means that there are almost no direct communications, uh, between peer to peer. Uh, everything is gathered around the, um, slack channels.

So each GU has a channel, each squad has a channel. Um, sometimes it’s like key and squad channels, and then you raise questions, uh, not in like direct, uh, messages. You raise question in the channel and, uh, pretty much, uh, regardless of who is currently online, someone can pick it up so you receive, uh, questions much faster.

And also, of course, it’s all of the knowledge that actually is gathering in one place. And you can then. Uh, use some of the participants of my research. They used the, the slack messages that they collected during the years. Um, they linked the language model, uh, to that one, [00:13:00] and then they built the internal knowledge base.

So currently it’s not like list of structured policies, it’s just a chat where you can ask, okay, how to submit vacation. Uh, I am like, uh, this, uh, I’m in this squad, and then like, how to submit vacation and you receive an answer from the chat, which is also like, uh, a perfect solution from my perspective because like when you read the policies and you know, you know, you need to go through the questionnaire.

It’s, everyone hates that at the beginning of their journey in any company. And this, uh, thing, uh, totally eliminates that one. Uh, also, uh, maturity of the team allows to, um, to shape your, to, to do more work, uh, uh, dedicated to your role. Meaning, for example, um, if you product manager. Then you can do some research, some, uh, play what if scenarios, AB testing.

So think about the strategy of the product rather than actually managing the teams. Uh, and uh, this is also common across all successful remote, [00:14:00] fully remote companies that I interviewed during my research. Uh, we also like document everything. Everything is is either in Google Doc, confluence, Figma, uh, some Bridges session, is there some, um, Vicky is, is there and that.

The AI with ability to actually generate automatically, let’s say, um, nodes based on the meetings, those kind of tools, it helps a lot also with this part. So it was always like that. But, um, uh, currently AI tool simplified this part of work because, uh, usually we didn’t put notes like meeting mind notes was never existed, uh, on the roster, but we put notes directly during the meeting.

Uh, it’s much easier, like everyone has access to some shared documents while I’m talking. Someone is putting notes, um, about what I’m talking about. When, uh, someone else is, uh, um, starts to talk, then I put notes for that person and it worked like that. And now AI is doing that, which is kinda, [00:15:00] um, also, which is the

Sergiy Korolov: thing.

Um, yeah, so I just wanted to add here, uh, quickly take the, the lead in this. Um, so originally the roots of rails were, were remote friendly. So it wasn’t like remote first. It was remote friendly. We had office in Kyiv and then we had a lot of folks who liked, uh, Ruben Rails. Uh, they were spread it across the countries.

So that was the, the foundation and all the processes started from there. Um, big, like majority of the company were engineers, like, I dunno, 35 engineers out of 40 people were engineers. Uh, so we had really, really thin layer. Yeah. And other five, uh, two of them were engineers as well, were, they were just managing the things.

So like, it’s mostly engineers company, but it was back then. Um, and so engineers love to automate things and the environments around themselves. Um, [00:16:00] so we did that that way as well. And, uh, kind of then the company was growing. New positions start to appear and we applied a lot of, uh, mechanics that we use in, uh, engineering world, uh, to the other positions, for instance, pair work.

Uh, so pair work in engineering world is something very kind of common, but uh, in the other spheres it’s not that common. And we started to practice it with product management, product design, operations, uh, recruitment and other things. And it appeared that it boosts teams a lot. So, um, again, talking about the remote friendly cooperation.

So you go to the call and you sit together, you work on something, it makes you productive, it makes your colleague productive. You share knowledge quickly. Uh, you motivate each other kind of, uh, you will not open, I don’t know, kind of the, the newsfeed and scroll it. Uh, when you sit together in pair, I [00:17:00] mean, that’s, that’s a real thing, right?

To kind of mm-hmm. What happened when the COVID times when people, uh, I, I’m not saying that there’s, this is the real reason why we started, but that’s a, some, uh, some additional benefits that we start to, to see, uh, that there are when you work in payers.

Um,

Sergiy Korolov: yeah. And, uh, when COVID started, uh, for us it was absolutely no change in the way we work.

For the other companies, uh, it was a drama, of course, uh, kind of even very, very famous companies, uh, IT companies. For them it was like no option to go and work from home. Um, so some of the smaller things that we use is usually I mentioned, so when we collaborate on something like let’s say we have, we’re starting new feature, big feature, or we start a new product, there is a Bridges concept, uh, framework that we use, uh, collaborating on the virtual whiteboards, uh, defining subject risks, [00:18:00] issues, benefits, solution variations, uh, kind of, uh, making researches and collect all this information on this whiteboard.

So whiteboard is the crucial part of our collaborative, uh, process. So if we talk about software development, there are smaller things like, uh, standup notes. Which team members, uh, usually, uh, write, uh, beforehand. So it’s very important. It’s, it, it sounds like a simple practice, but it’s very efficient. It allows you to speed up the stand up, uh, very much because like you do not see, like stay and think, alright, what have I done yesterday?

What did I plan for today? So you take your time beforehand, uh, you spend it, you make it structured, so it’s much easier for you to, to read it, uh, than, um, uh, uh, uh, I did, you know, so you, you, you make it really sharp and clean and also it, uh, gives a possibility. [00:19:00] For the others, uh, to see like, how productive are you, right?

Even retrospectively. So, uh, of course there are cases like we’re super productive. I would say our team is very productive, uh, sales, sales sufficient, but still there are cases like some, I dunno, burned outs happening, of course, kind of our organization is more than 200 people. So different things can happen, of course.

And, uh, kind of, you can retrospectively go through the documentation, go through the kind of, uh, standup notes, uh, like what was the progress. So it’s easy to kind of, so it’s, it’s actually a log right, of, uh, what has been done. So, and of course all those, uh, JIRA projects and other things and ai, uh, which you, uh, already mentioned before.

Galen Low: First of all, I mean, I a, I love all that. I love the sort of engineering DNA because some of it like really comes together, like you mentioned about the para programming and the documentation. Uh, yu I just love that you’re doing your dissertation on that. I think it’s [00:20:00] a hugely good point that like, I don’t know if we talk about that a lot, the fact that like, there aren’t that many organizations that are just co-located now, like hybrid and remote is, are the two sort of dominant options right now.

Um, and then as you’re saying that I’m like, like my, like, I don’t know, non-technical project manager brain is going like, Ooh, but like not having one-to-one conversations. Like, what about like, can we still say the same things we’d say to one another, but then I’m like, oh, but of course, like, it’s not like all of my conversations that I have one-on-one where I think I’m, you know, doing a good job of politicking for my projects are locked.

Like they’re locked the way they’re private. They’re not gonna be to the benefit of my team. And it’s not gonna be to the benefit of any AI tool. Like if you’re training models on some of this, like that’s all locked away. Uh, and now I’m just going through my head going like, oh my gosh, how many conversations, how much work do I do that is locked away that I can never actually benefit from, uh, either in a collaborative context or in an AI context.

And yeah. Anyways, it gave me pause. [00:21:00] Um, I’m just curious, you mentioned about the culture and the hiring. Like, does anyone come in and go, wow, this is really weird, like that we don’t have one-to-one dms that like, I only spend, you know, like less than an hour in meetings a day instead of, you know, being on the phone all the time.

Is that kind of like, is it ever a turn off? I was, uh, Julia,

Sergiy Korolov: sorry, before, before I start. Uh, so it is kind of not forbidden, but it’s highly recommended to kind of write everything in channels. Uh, again, because like when you leave the company, what happens with the information that has been, uh, shared, uh, in those channels?

What if some other team members have their kind of vision on what to, what should we say there? So kind of, it’s, it’s very important, like small thing again, but very, very efficient. Don’t write personal messages in Slack Forget about it. Like public channels or private channels. Uh, so make a group of two, three people, make it private, uh, but uh, write everything in channels.

Go ahead. Sorry. Not,

Galen Low: not prohibited. Get it

Julia Ryzhkova: [00:22:00] answer in your question. I was that kind of person. So when I joined Rose and I shared this experience, I was super insecure, I would say, because like, you know, usually when you’re a manager and you’re a good manager, you feel like you’re needed because you join and everything just stops working when you leave.

So it’s kinda like you are crucial part of the system and like, you know, two weeks, three weeks, uh, you became that part of the system. And then, okay, you can take one week vacation probably with your phone, probably with enabled emails. Mm-hmm. Um, but then at the same time, like you. Understand that if you will be gone, you can be replaced, but it’ll be hard.

And also it’s like you, you, you don’t feel that because you join the mature team and they actually do their stuff without you. And then you understand that you can take, like, I had a vacation for one month during this winter and they delivered a huge kind of things like, uh, nothing. I mean, for one month, 14 people were working without any management at all, and they managed, uh, to deliver um, about 10 plus releases.

So kinda. Uh, [00:23:00] outstanding things. And then you start to ask yourself, okay, what is my values then for the company, why they need me? And actually what you do as a manager, you just make processes and select things so that they will be working on to deliver more value. So they can actually do whatever. They can execute everything, but your value is actually makes them work on the proper things.

So it’s not about, you know, the difference between efficiency and effectiveness. Mm-hmm. So that’s kind of your value as a manager. And that was, uh, something that I realized only on the second or third months, um, during my trial with Rosa because everything was so uncommon. I would say that you need to actually switch your mindset to be a part of this culture.

Galen Low: I think that’s so like, it’s so thematically relevant, right? Because yes, there’s that manager ethos. You’re like, if I’m not in meetings all the time and telling people what to do, then you know, either things will fall apart. Or if things don’t fall apart, then I’m probably not adding value. But then there’s [00:24:00] that question of, well, how can I be adding value?

Like, I’m still here. What, what can I do? Um, and I think the AI question is that for a lot of folks, right? It’s like, okay, well it seems like I can step back and like AI is gonna, you know, create my meeting notes, it’s gonna like create my project briefs. It’s like writing user stories, it’s tasking the team.

It’s like checking in, it’s running standup. Like, am I out of a job? Or I guess the other question is like, well how can I add value? And I think, I don’t know, like, are you finding it similar in this, like, as some of your teams are adopting AI technology into the processes, um, are you finding that it’s like, okay, yeah.

Like the value I’m adding is sort of giving a framework for some of these processes to happen where we use ai, where we’re collaborating, you know, the ways that we are collaborating, how it’s changing our ways of working. Or is it also that moment where you’re like. I don’t know if I’m, you know, I, maybe I’m becoming less needed all of a sudden.

Sergiy Korolov: Can I briefly drop a metaphor here, uh, in this conversation? So I see that product managers, product managers at the moment, they’re [00:25:00] very similar to managers in soccer teams or like any other football teams. Uh, so is it possible the team will play a game without a coach? Absolutely. Kind of. They will go and play.

They may win something, even like top team may win something. They may lose something, but they also may lose this kind of coach as well. Right. And so there is this, um, thin layer of contribution that is kind of possible to. Ignore for a short period of time. Like Julia said that she went for a month to vacation and everything went well.

But if she would go for a kind of six months is then the processes would start to work differently. Like three months is probably. So po is, is, is, is the role That is very kind of common, I would say, or similar to what project managers, project managers do. They make this team play together, see inefficiencies, uh, [00:26:00] fix them, support them, uh, find, uh, any additional kind of resource, uh, to cover certain areas that are not covered by the other team members.

So that’s, that’s the main gen, main, main job I would say.

Galen Low: I really like that metaphor. Like, I like the sort of like the time aspect of it too. Like, yeah, you can play a couple games without your coach, but the coach is kind of like this, like overarching like storytelling team. You, you will for the team, not,

Sergiy Korolov: you will not win the season, right.

Uh, without coach you will not win. Uh, but a couple of games you can win.

Galen Low: That’s fair. That’s interesting. I, I, I’m wondering if I can, um, like move into like both of your experiences, like rolling out AI at Rails Wear. You already had this culture that was, you know, it started remote friendly. It’s got engineering, DNA, uh, you’re mostly remote first.

I mean, I would define it as that in terms of the ways that you work and then you start adding AI to the mix. Sergei, to your point, there’s probably a lot of folks who are, you know, technical, um, who may or may not have [00:27:00] been embracing ai. And then like you said, there’s like other cross-functional teams.

There’s the marketing teams, there’s, you know, the business side of things. Um, how have you found it to be, to like. To roll out sort of AI formally like governance and policies within that culture. Um, like has it been jarring? Did it kind of start without you? Um, yeah. How was it? How was it playing out like in the big picture?

I.

Sergiy Korolov: So I will start briefly, high, high level, and then Julia can tell more because she was one of those who, uh, actively pushed, uh, ai, uh, in, in engineering aspect, uh, uh, in her team specifically, right? So on. Again, uh, let’s think about, uh, elsewhere as the graph. And then this graph has inputs. Uh, every node has inputs, so all the else are nodes.

So they start to get inputs from the internet that AI is popular and so on. So they generated perception, uh, [00:28:00] how AI can kind of improve their lives and how I can put some threats, uh, on their position as well, because, uh, that aspect is very important, this psychological. Uh, there are a bunch of articles that you can find, uh, on the internet.

Pretty interesting researches. Um, so speaking about those threads, first of all is that a lot of, uh, you know, those influencers talk about that. Uh, AI will take a job, will take a job, will take a job, and kind of when you listen to that all the time, you kind of a bit stressed out and you start to think, you know, argue like naturally that maybe we shouldn’t be.

Maybe I can, you know, uh, postpone this introduction of AI into the company and then my job plus longer. Uh, but, um, so far I don’t think, well, of course I, um, I don’t, I can predict future and how fast the a GI will appear, but. So [00:29:00] far, AI, we consider as the very, very powerful tool that, uh, gives you a possibility, deliver better and more.

And, um, you know, few years ago we started with the content creation, with the marketing aspect. And you know what, uh, we, we have hired more content writers since Judge GPT appeared on the market. Uh, rather than let some content writers go, uh, kind of, we just started to generate more content and more, more quality, better quality content.

Uh, so yeah, our content writers, they’re responsible for creating story and kind of, uh, collecting interesting information around the company. And then they generate, uh, very, very good, interesting and useful articles. Um, so yeah, and the, that, that’s content writing that, that started, uh, with, uh, of course we started to use all those instruments, uh, that Julia mentioned already.[00:30:00]

Uh, low note takers. Uh, we had this culture taking those during the code, uh, call, but then you cannot take them as, as great as, uh, AI can do right now. Right? So there are all the aspects. So there are summaries, and that’s, that’s really fantastic thing. So you may skip some conversation and then just read the brief summary and understand if you need to go deeper so it reduces the time,

Julia Ryzhkova: or even ask ai, read the, the notes, and then provide you only summary ahead, which I usually do.

Sergiy Korolov: While, while driving a car. Right. So that, that’s a very fantastic thing. And, uh, yeah, the, the biggest change was of course, uh, with the engineering parts. Uh, so for quite some time there was a hype around co why vibe coding, uh, right. But, uh, two years ago, let’s say even a year ago, the quality of the code, uh, was, uh, kind of so, so [00:31:00] kind of do, our engineers started to experiment with AI tools like copilot and, uh, later course or, uh.

Pretty fast 2080 period in the market. But of course, when you do experiment and you get some bad results, uh, then it doesn’t mean that tomorrow you will go and test again. So there were like experiments then, uh, kind of overall conversation in our Slack channels that, you know, we tried. It doesn’t give, uh, so much, uh, kind of, uh, benefits.

Uh, so nothing great. Then they’ll work, uh, different. Concerns regarding, uh, the policies. So sometimes ago there was no clear policy. If AI will, uh, learn, uh, educate their models based on your code, uh, based, or it’ll not, uh, but later, kind of the, those issues has been addressed and the kind of, when you purchase certain plans, in course copilot, then, uh, they guarantee [00:32:00] you that they do not educate their, uh, language models on Euro code base.

Yeah. It stays kind of locally. Um, yeah. So, but then, uh, again, different parts, different teams, uh, different individuals, uh, start to experiment with, uh, ai. There was some progress, but in this specific case, there was a need in push from the high level, like from C level, uh, to, uh. To make it happen and kind of start to apply this AI, uh, usage across the teams again, because there was some skeptics, uh, historical skeptics, uh, some probably like resistance or not acceptance that, you know it again, when you constantly hear that.

Uh, I, I love that the YouTube, uh, video that has the name I’ve built Calendly in 15 minutes, you know, and then you have a Calendly team, which is working for eight [00:33:00] years, uh, or more than eight years already on this product. And, and kind of you see such titles, uh, and it is a bit scary, uh, but that, that’s, at the moment it’s still a hype kind of, you will not be able Calendly in 15 minutes.

That’s, that’s a bullshit. Um, so anyway, um, I. We started to, to search for like one or two advocates, uh, within the team, uh, in every team. And then, uh, work with them closely asking them to collect the cases and then make presentation to the rest of the team, uh, showing the best case scenarios, how to use, where the contribution of AI is.

Very good. And, uh, uh, I have a list of those contributions. We can talk about them later. I guess. Julia, maybe you can adhere.

Julia Ryzhkova: Mm. Actually I was thinking about, uh, maybe our audience would be more interested into how we leverage AI specifically, let’s say in areas, [00:34:00] uh, according to handbook, uh, because it’s like more, more moral relative, I would say about the project management.

So of course the, uh, more, the easiest one is, uh, the communication because, uh, like any questions that you can get from, um, customer, which usually took you. Ages to craft the proper feedback. You know, like some comparison, like why, why we should do that, why we don’t do that. Uh, some why we are scheduled two release and all of those kind of things is currently is done like that.

Uh, based on just so I enable a voice, uh, in charge, pity, then I tell about the problem and then I receive a perfectly crafted Slack message, confluence article about some details and those kind of things. And it actually takes minutes, uh, when I am driving or going to the office, um, rather than when I sit like in, you know, in two, three hours, you are shaping the like, wonder wonderful results.

So communication. Mm-hmm. But I would say that this is probably obvious and everyone are doing already that, um, in their day-to-day on day-to-day basis, [00:35:00] um, uh, from these, uh, let’s say integration management, it’s uh, like currently that’s probably the biggest challenge we currently have. So in order to integrate all of the things, it’s still on us.

Uh, it’s still on the, uh, teams themself. Um, uh, so far we have this vision, but we are currently only working on, uh, executing it to set up AI in a way that each team, uh, in the processor uses the previous artifacts, and they all are managed within the same MCP. Um, however, it’s, uh, only division right now.

Like, uh, for example, we have the kickoff session, then we do the bridges, then we make a screenshot of this, uh, brainstorming session, and we have, uh, let’s say some prototype, uh, in, uh, like. So far without any style because all of the application that actually generates prototypes within ours, they can generate it only once.

You can change it, because everywhere everything is broken. Um, however, like the first one, uh, you get within ours and, uh, this is something you can already talk about as a [00:36:00] corridor testing and things like that. Then you do Figma. But then Figma is being used for producing requirements. I currently do that, so I make the adjusting screenshot.

I have a separate project, um, in, uh, change GPT, uh, one project for making data analysis. And it has, um, as a files of the project, uh, my structures, the database, let’s say, so that I, I can easily like query. And ask it, uh, to, to prepare some queries for ad hoc analysis and the separate project for creating user stories and those kind of things.

Um, where also all the context is already, uh, applied. And then I just make a screenshot and I say, okay, this is what we want to change. Give user story and it gives user story. And of course we do the same to get their, uh, test cases and acceptance criteria that’s, um, exactly like that. And, um, like what, what we want to achieve so that it will not be manually.

Like I open Figma, make a screenshot, put it to chat. Then I got, uh, those things. I put it to Jira. So [00:37:00] currently more and more MCP, um, um, agents are, uh, being released from different tools and we want to integrate all of them. So the, the vision is that it’ll work out together, but we are so far not there yet, let’s say.

So. Integration. Yeah. And we’re

Sergiy Korolov: going to release MCP in COUPLER very soon, so you will be able to query your. CM your marketing tools, your I know, a TS tools and financial tools, uh, with the natural language, uh, kind of cooperate with your data. Nice.

Galen Low: It’s a, it’s a really good tie. Like someone earlier was saying, you know, I’m being asked to do ai, but like, none of it’s playing well together.

Uh, and we had a session, uh, a few weeks back about like, yeah, I feel like I’m just a copy and paster right now pasting from one AI tool to another. Um, but I really like, uh, like really, I really like that, um, uh, that, that, that use of like projects, for example, right, to train like a project, uh, in chat JPD, uh, to.

Serve a purpose that already has the [00:38:00] context that not really kind of understands how to do the job, to kinda like just shorten that time and, gosh, relatable, that whole like blank page syndrome where like, Hey, I’m gonna sit down, I’m gonna spend three or four hours putting this, you know, document together versus dictating in your car.

And if I’m understanding correctly, dictating in your car and getting it into Slack and Confluence, like in the same go that’s, I should have, I should have titled this work from car.

Julia Ryzhkova: Uh, but the, like, if you have this, um, you, you know, I always, uh, admired like in Ironman, those like Jarvis. And, uh, then kinda, we are almost there here with our current Jarvis.

Uh, so mm-hmm. It, it’s like, I, I talk to AI a lot, uh, because it’s much faster than just put in your, um, prompts. So usually what I do, I ask verbally. I give a context, and then I ask verbally to prepare a prompt for it, and then it try a prompt for me. And then I start a new chat and I use it prompt to actually properly prompt it.

And, uh, like, uh, for example, it [00:39:00] took me, I don’t know, we, I had, uh, uh, also on my dissertation, I needed to prepare at, um, presentation. And it took about 15 minutes for me because, like I talked to, to chatGPT it created a prompt for gamma up and then gamma up, uh, generated, uh, perfectly, um, perfectly structured presentation and everything.

The only, uh, downside was that the HR manager was a girl with a bird. It’s still like that in the eye, but it was diversity. Diversity, yeah. So it’s kinda, um, and that was the only fix that I did. So, 15 minutes and you actually have a like, uh, uh, it was even including some infographics. Uh, and like I use AI a lot, however, I always recheck the results and that’s the main thing.

So, um, what I noticed, we also use AI in hiring process, of course, but the good thing is that we have a specifically defined criteria, let’s say, for test tasks. So that we, when we evaluated with [00:40:00] ai, it actually can give some feedback. It’s not just, uh, is it good or not? We have a criteria and then it works.

Um, how, and also of course, uh, as a part of our hire process, we have payer interview. Uh, Serge mentioned that we have a lot of payer work. That is why we do per session together with candidate to person from our side and candidate. We do some things like, uh, data analysis, let’s say, or, uh, some researchers, those kind of things.

And, uh, they tend to use AI and we’re allowed to do that because it’s like, it’s our nowadays. However, we al always ask it, it doesn’t help people usually if they, they are not like on a good level, they can’t leverage this too. So they ask questions, okay. Uh, like, here’s a data set. What insights can you give me?

And it gives not a good insight. Some, some of the ideas. And then I say, can you check it like on your own? Just check that it is true or not. Or for example, Gemini, it, uh, created a like sales dynamic. And I, I’m looking, so I know this data, it [00:41:00] shouldn’t look like that. And then I understood that it sorted months alphabetically, April, March, may, that, uh, April, December, sorry, March May.

And those kind, and I’m like, I, I’m looking at that and I noticed it, candidate did not. Mm-hmm. And it kinda, so, uh, this is only the two. So you still need to use your brain to check the mm-hmm. Uh, to check the result. And that is why I, I’m not, uh, like, um. Uh, concerned that I will, uh, I will be losing my value, let’s say I can just do mm-hmm.

More things. Uh, I can write those security queries on my own, but it’s much faster for me to say, I need to understand like, how many, uh, do we have more people who sign up from this channel? Like, uh, who use this feature in the recent one compared to the previous one? And then I have a scale I just executed.

I can write it on my own. Mm-hmm. But then why do I need to do it?

Galen Low: I love that it’s in the interview process, this sort of pair session to kind of evaluate team members, prospective team members’, [00:42:00] ability to evaluate AI output. Right. It becomes, can you see what’s missing, what’s wrong, what the gaps are not.

Can you write a SQL query? Like that’s,

Julia Ryzhkova: we don’t, we don’t force it, it’s just that we say you can use whatever you want. Mm-hmm. So let’s mention your world. You need to do a decision and they say, can you use, uh, change GPT. Go ahead. But we also put in efforts to, to shape our things that we put in our hiring process in a way that you can’t get, uh, good enough result with ai.

Mm-hmm. We tried.

Mm, mm-hmm. We

Julia Ryzhkova: know that you need to, so, you know, like it, it’s the question that contains, uh, the answer. So if you prompt it in a way that you know what you’re doing, then you can get the result. But if you will just, uh, ask dumb questions, it’ll not give you anything. And we know about that.

Galen Low: That’s super interesting. I love that. Um, I wanted to move to audience questions shortly here. I did wanna come back to one thing, Sergey, that you had said, um, because I saw it in the chat. You know, there’s a lot of folks here who are like, I’m being asked to use ai. No one’s telling me [00:43:00] how. Um, you know, my business owners are trying to replace me with agents.

Um. Something that you and I talked about Sergei, about like, uh, like you’re quite a technical person. You have a technical background. You mentioned like YouTube videos, like, oh, how we built Calendly in 15 minutes. Uh, and you said something about like, yeah, that’s kind of like, that’s hype. Um, but there’s probably a whole bunch of like non-technical business leaders who are thinking this.

They’re trying to create an advantage for their business. Um, and they’re trying to like, filter out all the noise, but sometimes it seems pretty feasible to be like, okay, well yeah, I can just, you know, get 300 agents and you know, like, uh, have my staff, um, and double my profit and apparently I can do it in 15 minutes.

Like, what advice do you have for people looking at this? Like how can they evaluate if they’re not really from a technical background, how can they evaluate? What is sort of like the BS versus what is actually something they should be taking? I guess

Sergiy Korolov: there, there [00:44:00] must be some community, uh, kind of like open source.

Usually engineers do open source when they try to resolve some problem, right? They, they need the, I dunno, operational system. They create Linux and uh, like, it’s kind of a joke, but not a joke. Uh, so you need some foundation. So, uh, anyone can have at least some arguments to non-technical founders, co-founders who blindly believe that AI already has some general, uh, AI skills.

So generative, uh, ai. Uh, so it’s, it’s not generative yet. Uh, kind of, and, uh, I mean. So far what they say is that LLM will not bring you to JAJI technology. So there must be a different architecture there. Uh, but, uh, we, we see, we see this a lot actually. So it’s, it’s so easy to get into that trap. You, you watch those videos with that YouTube full [00:45:00] of that really, you can do this and this, you can, I dunno, build, uh, a tool, uh, which will replace, uh, 10 content writers and kind of, and you will generate content with this tool.

And for a certain period of time, it worked even with the content. But then Google Bond, uh, those, uh, websites, uh, fully generated with AI content and that’s, it, kind of, that strategy is over, but no one told, tells you about this later on. They, they will not, uh, deliver you a message that, hey, it, it fucked up.

So kind of, it worked, but it doesn’t anymore. Mm-hmm.

So

Sergiy Korolov: they just, uh, show you the kind of, uh. Positive side that, oh, you can do this and this and this, and on. I, I believe that’s a big problem, honestly, that, uh, uh, there is some, there are some illusions, uh, of co-founders and founders because, uh, someone, uh, CEO of AI company draw a [00:46:00] picture of the future, which is not a current, uh, it is not a current state of AI that will give you such benefits, but they draw you, you know, the future, how things will look like, like Ellen drives, uh, like sell you that, uh, they’re gonna be a self-driving car.

And he sells this for many years. Uh, kind of, but still, it’s not that, uh, kind of that autonomous, I mean, it is autonomous of course, but when it will make a mistake, it’ll be enough. One mistake, uh, a year that you will not survive it.

Uh,

Sergiy Korolov: so. But my point here that CEOs of, uh, big companies of who run ai, they sell you the future, which is not the current and the other non-technical CEOs, they, uh, buy it and try to squeeze from their teams, uh, kind of the results that the CEO of those company.

And it’s messy. But this always happens when new technology appear kind of, uh, there [00:47:00] is some, it’s a really breaking change, right? So it changes the, in industry for sure, changes the industry and the way how communities leave and work and cooperate. Um, but um, yeah, sometimes there is too much hype or, and, uh, yeah, it is very good by the way that, uh, apple released, uh, this white paper, right, uh, analyzing the real capabilities of ai and at least it’s some argument that you can right now put on the table in front of someone who already believe in some, uh, AI capabilities of current ai.

Galen Low: Yeah, I love that. Like the education of where things are actually at versus where they should be in the future. And this notion that like, I. You probably do need a human there to kind of navigate things that change, because you’re right, no one’s posting a YouTube video that says, remember that thing I said that you could replace, you know, 15 staff with one agent.

Yeah. That’s [00:48:00] changed now. Like no one’s making. Yeah. My business has bankrupt,

Sergiy Korolov: so

Galen Low: yeah. Oops. Uh, I like that. Um, we will try and link that white paper that you mentioned, uh, Sergei in there. Uh, I wanted to move into audience questions. Um, I, I, you know, we’ve got about eight minutes left. Um, so, uh, just before we get into q and a, uh, and just before folks start leaving to their next meeting, um, I just wanna say thank you for joining here today.

Uh, if you enjoy what you are experiencing here. Uh, today we’ve got a few more events coming up in June. Love to see you there. First one is about, um, how everyone is a pm Now maybe, uh, we’re gonna touch on some core PM skills. We’re gonna talk about the PMP and what it, uh, whether it’s relevant to folks who don’t see themselves as, uh, project managers.

Uh, Michael’s gonna throw the link to RSVP, uh, in the chat. Uh, and then the second session we’re gonna dive into is how you can use your PM skills to negotiate your next raise. So we’re gonna be talking salary. We’re gonna be talking about some of the trends that we uncovered in our recent DPM salary report.

Um, and it’s gonna be a lot of, uh, it’s a lot of fun, highly relevant. [00:49:00] Uh, Michael’s gonna post the link, uh, to that one in the chat as well. Um, last thing, we love feedback. Um, we wanna be able to improve our product, which is these events. Um, so, uh, if you could just take a moment, um, to take our survey, Michael’s gonna post that link as well.

Um, be raw, be unfiltered, be honest. Um, this is the way that we improve how we do things here at the digital project manager. Alright, um, let’s, uh, let’s get to questions. Um, I, I wanted to dive into this one. Uh, I know Julia, you’ve been answering questions in the chat, so I’ll try not to like, um, cover over those as well.

Um, but that’s, that’s Julia’s

Sergiy Korolov: multitasking, right? I was gonna say, yeah. This is her superpower.

Galen Low: She could have done this in the car just dictating. Yeah.

Sergiy Korolov: Yeah. She would probably, she’s probably driving in the car. It’s, you know, it’s a background like green background. She’s actually

Galen Low: an agent. Yeah. Generator space.

Yeah. Yeah, yeah. That’s my, I love it. Uh, I think this is like, quite relevant. Um, I’m gonna jump on this question from Jamie. Uh, uh, the question is like, how does relying on AI impact your [00:50:00] reputation as a reliable source? So if you use AI a lot, um, like could you fall into the trap of like, actually not using your brain?

So actually like the opposite of what you’re saying, where it’s like, okay, actually, um, I. And Jamie, tell me if I’m, I’ve got this right, but this sort of idea that you’re like less valuable, your, your perceived value actually becomes less because you’re using AI like in the workplace. And instead of a culture of saying, okay, well that’s great, you’re, you know, let’s find a way to level you up.

It actually starts negatively impacting your reputation as, oh, that person who just turns to ai, AI all the time and, and turns off their brain.

Julia Ryzhkova: There was a research about that one. Um, they, they had two, two groups. One, uh, the three actually, one is controlled, one, uh, it was forbidden to use the at all. Uh, they asked them to learn some python or something like that.

And then, um, those who used the, who were allowed to use it to ask questions to actually learn, they performed much better. Those [00:51:00] who were allowed to actually ask questions and then copy paste the result, they, uh, didn’t learn a thing. And then the third one, which is control one, uh, was the, in the middle of those two.

So if you use AI to actually learn things, for example, I’m working on, as Sigge mentioned, currently on Trade Zero. Uh, trade Zero. And it has, uh, a lot of context, trading context. It’s a really hard dom knowledge. So currently I’m learning, uh, uh, trading. Mm-hmm. Uh, and I actually use a lot of. Uh, and also like QA team engineers, they all need to understand how trading works.

We all, uh, are back testing. We all are, um, uh, trading a bit, uh, in order to understand what we are working with. It’s like, uh, uh, for traders by traders. However, still we use AI a lot because it has all those domain knowledge. Uh, and I ask things okay. From your perspective, for example, what is the most, uh, um, like type of instrument people are trading on, uh, by bit, uh, exchange?

Because I don’t know that, and AI does. So I use, [00:52:00] I ask questions to learn more, and that is why I believe that I would be in that first group. You know, the one will actually be, that is how you need to use it. So I’m not afraid about the reputation because it’s only the tool and it makes me more, um, effective.

However, I’m still responsible for making me, uh, sorry, efficient. However, it’s my responsibility to make me effective so that I will ask a proper question.

Galen Low: I love that example, like, talk about value. I’m gonna learn about trading so I can make a better product for traders. Like, amazing. Sorry. And,

Sergiy Korolov: and quick, uh, quick, uh, example here.

So it’s not the first time when, um, industries, uh, receive some powerful tool that changes completely the productivity. So I want to remind you that the engineers like, I dunno, 70 years ago they had those huge machines that took like the whole, like the, the, the whole flow floor. And, uh, they used, uh, paper cards to program them, and then they had terminals.

But then, uh, assembler appeared and it, uh, [00:53:00] was, uh, one way to communicate with machines, but then kind of, uh, higher level, uh, programming languages like Turbo Pascal period. And they be, they made them much more productive, but then rapid, uh. Development frameworks appeared like Delphi and one engineer was able to build a tool that before, uh, 20 engineers was not able to build within the such period of time.

And, uh, kind of, if you will look around, kind of, there is a constant progress. Uh, this open source community that generates frameworks like, uh, Ruben Bras, uh, which has the millions of lines of codes that already prepare for you. And you can generate a blog without AI within five minutes. Uh, you can set it up, right?

So it was even before ai, you were able to build a lot of nice things, uh, with the minimal effort, but then you still need to be a professional to understand how to tweak it further or how to move it further, further. [00:54:00] So same thing here and across the other professional professions, uh, that you receive the, you, you have a powerful tool, like, but you need to know how to use it.

And in the end, it’s your responsibility, how you treat the results, the outcome of the, uh, AI tool.

Galen Low: It’s a, usually, usually good point. It’s just another new tool in some ways so far.

Sergiy Korolov: Yeah, of course. When a GI appear, then we, we, we have fucked.

Galen Low: Yeah. Yeah. Then we have the session again and say, that’s why I learned trading.

Sergiy Korolov: Yeah. Yeah. There you go. No, but there will, there won’t be any need to like, to learn anything trading because it’ll trade themselves, uh, better than any people, any human.

Galen Low: Uh, I, I mean, you know, in the meantime, and actually, actually one of the

Julia Ryzhkova: arguments like, uh, we can use in communications with, uh, uh, customers, right?

If they say like, why, why you can’t replace your engineer with ai. You say, okay, you’re doing, let’s say pharmacy, why [00:55:00] we can’t replace pharmacists, why we can’t replace doctors. So it’s kinda, it’s, it is just not there yet, let’s say. Mm-hmm. Um, however. We can use it. I see. Like what elements of PM can be replaced by ai?

Yeah. I would say that one of them can be actually currently fully replaced by ai, but I mentioned some of them that can be actually improved, uh, the performance of which, uh, we can leverage it. Uh, like risk management for example, for example, you know, those kinda standard risk management for the project.

And actually you, you are thinking about which risks, uh, specifically are for this project. So AI generates that, uh. Pretty fast and, uh, amazing. I would say just based on some incoming, the documents, uh, and, um, the preamble of your project scope document and,

Sergiy Korolov: you know, and one, one more thing, which is very important.

So product management, project management is about people managing people, uh, not only tasks, right? In Jira. And so far, like I don’t believe that any [00:56:00] person will execute the order of, uh, judge GPT in like, uh, if it’ll say to do something for him. So kind of, uh, this role is not so far, it’s so far from being possible to replace kind of managing people.

So they’re gonna be managed by people when they are gonna be managed by robots. Then we have, uh, I don’t know, Terminator thing.

Julia Ryzhkova: Yeah, that was also one of the questions, right? How, how AI is, uh, deal with people. It does not, and actually the, it’s a cha the biggest challenge is about insecurity of the people.

So, uh, when, when we lead the team and we enable ai, so I, Serge mentioned I was pushing, uh, like we need to leverage AI because it can help us, it can improve performance. Uh, however, one of the things that I dealt with is exactly those things. Like, am I being replaced by ai? And then no guys, it’s kinda, I, I, I wouldn’t do it to myself and I use it a lot.

So let’s just, uh, use it as a tool.

Galen Low: That’s great. Actually, you know what, honestly, maybe a really good place to leave it. [00:57:00] Right? And in terms of the fact that we are integrating technology with people who are collaborating together. People are still important. We can use AI to learn. We can use AI to be more efficient work from our cars.

There is integrations that still need to be, you know, thought through. So we’re not copying and pasting between tools and feel like that’s our value. There’s an opportunity for us to elevate our value, to do more, to help our people, to get people more comfortable with how we progress, um, and how we do our jobs even better.

So maybe I will just, uh, leave it there. I know I’ve taken us a minute over time, um, but I just wanted to say thank you to everyone in the audience who stuck around. Thank you for being here. We appreciate you. Um, if you wouldn’t mind taking a few seconds to fill out our feedback survey, uh, we’ve posted in the chat, uh, at some point.

Uh, I would love to know what you thought of today’s session. Uh, and of course big thank you to Yulia and Sergei. Thanks so much for volunteering your time to share what you’re doing. I think you’re doing amazing things at Rails Wear. I’ve been fanboying about you for years. Um, I think I learned a lot today.

[00:58:00] Um, and Julia, we’ll have to have you back to like read your dissertation to us. Sounds super interesting. Great.

Sergiy Korolov: Thank you guys. Bye. Thank you for having us and share the feedback with us as well. Uh, so it’s very important for us. Absolutely. Thank Awesome. Thank you.

Galen Low: Thanks again. Have a great rest of your week.

Sergiy Korolov: Thank you. Bye-bye.

​