Transcript: Bringing Generative AI into Product: Shay Shitrit of Augury

Eyal David: [00:00] Hi friends, this is Eyal. You've reached Product Builder, and here you can listen to the most interesting product conversations. My journey began in 2012, when I learned what product is at my own startup. Since then I've worked in the industry, and since 2017 my company has been consulting for a variety of companies. My drive is to help all of us improve as product people. And before we start, if you're enjoying listening to Product Builder, please rate us five stars on Spotify. It gives me great feedback that the podcast is giving you value, and it would make me very happy. So come on, let's start! Hey Shay, how's it going? It's such a pleasure to be here, what's up? A delight to have you here. What's up, what's up, but it's a pleasure to have you here. Joining us is Shay from Augury, product manager, right? Right. Coming on the recommendation of Shuki, from one of the previous episodes.

Shay Shitrit: [00:49] The great Shuki. Shuki, the greatest of all. Tell us a bit about what your day-to-day looks like and what you do. Okay, so I actually work at a company called Augury, I've been a product manager for the last two years. Maybe tell us a bit about Augury? So I joined Augury almost three and a half years ago, in a Customer Success role. After I'd had a startup, a failed startup, but a great roller coaster. I joined Augury in a Customer Success role, a new place for me, into the worlds of machine health and what the company does. After half a year I actually got to... to lead the group and recruit a global team, to build work processes.

Eyal David: [01:31] And when I had the opportunity to move to product, I jumped on it. Wow, so you have to expand on that later, but let's just say in two words that you really came from the traditional world, from Unilever, all the team-building and all that, it wasn't something new to you before. So okay, you've been there two years, amazing.

Shay Shitrit: [01:45] What are you doing day-to-day now? So I'm part of the fleet responsible for the customer experience in the product. My squad focuses on the worlds of the facility. That is, our personas are technicians on the production floor, maintenance managers, plant managers, who basically want to know the bottom line: what the state of their machines is, whether there's a problem and what they need to do. And this is a world that, as mentioned, wasn't new to you beforehand, right? At Unilever, that's the thing. Right, exactly. So after my studies — I studied industrial engineering and management at the Technion — I had to choose, basically, whether I'd go. To a startup, the tech areas that really appealed to me, or go to traditional industry, the manufacturing industry, which really fascinated me at the time. I really think manufacturing is magic. And I went to Unilever, into a management-trainee program, a cool program where basically every three to six months you switch roles in the company, and see the business from every angle. So I was in project management in manufacturing, moved to finance, moved to sales a bit, like I drove around in a car and tried to sell Unilever products, and in the end I led the planning and operations of the Haifa site.

Eyal David: [02:56] It's a bit like opening a McDonald's franchise, right? You have to go through the whole chain. And you said the sentence where you talked about manufacturing being magic.

Shay Shitrit: [03:05] Why? Can you elaborate on that a bit? The processes of manufacturing, of turning raw materials into packaging, something where you see a sack of salt here, you see a cubic meter of water there, and in the end finished products come out, reaching billions of people every day, this chain is insane. When you're inside a manufacturing plant, the chaos, the noise, the mess, when you experience it there, you say, how the hell does this work? But in the end it works, it even works efficiently sometimes, and it's just magical. It's magical, to take some process that someone did by hand until now, and basically break it down into tons and tons of mini-processes, that someone else, that lots of people do, each one doing their little bit, and in the end it all converges into some crazy machine,

Eyal David: [03:49] that produces something at scale. Yeah, I really connect with that too. So okay, you know the persona, you joined the company, you also came in wearing the hat of basically talking to customers, being involved in all the customer-facing stuff, and you basically arrived at product very mature, so that's very cool, and basically you're now dealing with experience, right? Right. So Shay, what does the experience actually look like, let's think about it for a worker on the production floor, with and without Augury? Okay, great. So without Augury.

Shay Shitrit: [04:20] Without Augury, basically uncertainty, and an inability to plan. If a maintenance worker on the production floor is responsible for the machines working properly, and making sure everything works as it should, and ultimately produces a finished product, he has a lot of downtime, a lot of failures. A machine starts working, suddenly something breaks, a failure that develops over months, and they don't know. They work in what's called a route-based configuration. Route-based means that either in-house technicians, or contractors, external subcontractors, basically go between all the machines, with some vibration meter, and in the end submit some report. The machines are fine, the machines aren't fine. And this is a very wasteful method, inefficient, and on the side, at best once a quarter. You say, okay. So that's how they work without Augury. And then Augury comes along, a game changer, and says: with sensors that we developed, installations that we do, we basically prevent Unplanned Downtime. That is, there are no unexpected failures. We monitor the machines, and sample them every hour, and we know how to say whether failures are developing months in advance, what type of failures, and what we recommend doing.

Eyal David: [05:38] It changes, it changes the whole rules of the game, it changes their whole way of working. Amazing. And I remember, when Shuki was a guest here, and we'll stop mentioning Shuki — so when he was a guest, there really weren't any competitors,

Shay Shitrit: [05:51] chasing after you. What does it look like today? So Augury is still the market leader. They were the first to build the category, and shaped it into what it is today, and we're still leading there, but the competition has grown. Competition has grown such that, certain geographic areas, where it's easier to penetrate with local products, and basically a war around differentiation, and what we do differently from the competitors. Amazing, which territories can you elaborate on a bit? So we work almost everywhere in the world, now we're also entering China, so we're really all over the place, the main market is really in North America, the US, Canada,

Eyal David: [06:38] because on the face of it it's also China, right? It's just probably really interesting, because all the factories in the world are in China, basically, right?

Shay Shitrit: [06:44] Right, right, that's why it's such a next target, that's very hard to enter, because there's probably a local player there, and regulation,

Eyal David: [06:52] and barriers. So basically you're saying, you create barriers — so what do these barriers look like? Have you already seen a bit?

Shay Shitrit: [06:59] So one of the barriers, I think, is the differentiation it gives us — we focus very much, and this is Yossi's squad, focuses very much on increasing our value to the customer. And basically, as part of that value, in our strategy we leverage all this enormous data that we have over the years about the customer's machines, to give him, in the end, insights that, well, others can't give so easily. Now, specifically, we focus on the worlds of implementing Generative AI inside the platform, which is — think for a second, as a maintenance person, you can talk to your machine. Insane. Or to your machines. Or even more than that, with all your machines. What's the problem here? What's developing here? Have I had similar problems in other machines? Who solved these problems? Who's the knowledge expert in my organization for problems of bearing wear?

Eyal David: [07:50] It's mind-blowing. How does it work? Is it really going to be verbal? Like with speech? Or how is it going to be? So in the first stage,

Shay Shitrit: [07:57] we're basically — it'll be text-based, and after that, there's no limit to what can be done. And is it basically your own engine? What's behind it? Can you elaborate here. So we use external models, but the core here is our data.

Eyal David: [08:11] Your layer. Exactly, it's our data, which is enormous. It's basically enormous. And what also intrigues me — you don't have to say, just — it's really interesting what the UX looks like for manufacturing people, when on top of everything you're now adding Generative AI into the mix, what does all this fusion look like?

Shay Shitrit: [08:30] At first, very very simple. That is, if you enter... it has to be simple, and that's what I'm striving for. Exactly, really really simple, very clear. It's, one sec, AI, predefined questions, plus your ability to write freely, but we understand what the things that most interest our customers are, our users on specific pages. And then with each page you reach, we know how to tailor predefined questions for you, in that context.

Eyal David: [08:57] Interesting. And how much openness is there really on the users' side to things like Generative AI? So we don't know yet. It's really sexy. But you basically did some process, you as a customer person, right, that's where you come from,

Shay Shitrit: [09:12] so you basically tested the matter — what did it look like for you? In our feeling, and in the MVP we presented to customers, it seems that it's very simple, straightforward, and there's very very high usage there. It also matches what we see — that customers are already asking our experts anyway, in the threads that exist, we're just saving them all of this. Think for a second, now with our launch of the mobile app, where the maintenance people are next to a machine, and can kind of understand what you're picking up live, and give it a solution, so now they can also talk and get some vibration analyst 24-7, right next to them,

Eyal David: [09:53] trained on all their data — that's something they never even dreamed of. Really, and that's also one leg, and the other leg is that it also helps you become more efficient as a company, basically. Amazing. Okay, so let's talk for a second about product skills. What, in your view, are the most important product skills,

Shay Shitrit: [10:10] that a product manager can bring with them? Great question. I think the first thing is high EQ. Very high emotional intelligence. I think at the heart of it, a product manager needs that aspect of empathy, the ability to create very good, very strong connections, to see things from others' points of view — that's something that serves a product manager completely. So that's the first thing. The second thing, I'd say, is leadership. It's to bring — not necessarily by virtue of authority, but out of some other understanding. Within the squad you're the person who brings the broadest take, the deepest, and you... everyone has an opinion. Everyone has an opinion, but how do you bring everyone to act around the thing, and this idea that you think — and if you look at the company overall, then you're basically the domain expert for this world. And when you see that at the edge, the users you focus on experience some pain that comes from other worlds too, then you need to come and kind of do alignment within the organization, and influence other squads and other fleets, to fix the situation. The last thing I'd say is a just-do-it kind of approach. A focus on execution, to progress fast, to avoid analysis paralysis, and to run with things. I think I'll say here, Augury, in its approach — and this comes from our founders — is very much about innovation, about power, about driving forward. You know, from the top, through our leadership, the VP Yedidya, who say: do, break, learn, improve. And it's like, we say this everywhere. No VP interferes. With determination. Go out fast, learn, and fly off. So that's the third thing I'd say. Just do, start with something. As long as the general azimuth is good, it'll be fine. It just needs to move fast.

Eyal David: [12:16] I'm kind of hearing you, and I'm thinking about two things. One, how important the first two things you said are at the company where you work, you described to me earlier — you basically work... what are your conditions regarding the office? Sum it up for a sec? Yeah, so we work mainly,

Shay Shitrit: [12:32] remotely. Once a week, twice a week, we come to the office. In addition, there's a very very good benefit, of a worthwhile Sunday — meaning four days a week, once every two weeks. And in general, Sunday is a day to catch up with things, and kind of take a breath of air for a second, and prepare for the coming week. So it's really fun, but I hear,

Eyal David: [12:56] like, from the outside, I say, wow, how efficient you have to be in these four days, and how important it is, basically, both the communication, and the EQ you mentioned, and the efficiency, and the just-do-it — like, everything, interesting. And your second question, and really this is probably on steroids at your place — the second thing I thought about is, since you arrived at Augury, which of these three do you think — because I'm guessing it also describes you, things you'd want to see in others — which muscle do you think you've trained the most in this period?

Shay Shitrit: [13:24] That you've improved the most in? So it's probably at Augury, and in general, in general throughout my career, so I'd say, the first thing: I really love connecting with people, I really love people, I really love relationships, building, understanding, and how to help, understanding motivations, that's my fun, that's my side, and it turns out I'm also good at it, so it's totally — if I were really one thing, then it's this.

Eyal David: [13:53] Nice, so the thing you went toward is to bring Gen AI into the product, and to look at the users — you, yes. Oh, cool, and what, did you really start touching this kind of during the conversation? What are your challenges today? So, if I look for a second at my world,

Shay Shitrit: [14:09] in facility engagement, that's what we call it, so we look on one hand at value — how I can give more value to the customer, in order to improve adoption and engagement. The second thing is basically how to reduce friction. Now, if we look for a second, at our users, this isn't the classic SaaS world of tech-savvy users. These are people who are on the production floor, who maybe don't have access to a computer, who maybe don't even have a smartphone, and if they have a smartphone, maybe they don't have internet, sometimes it conflicts with safety — that they shouldn't have phones next to working machines. And then you try to think, how can I make the right information accessible, at the right time to the right people. And then you start getting into places of, okay, so I have the platform, and I have the mobile app, and then I have SMS Engagement, Email Engagement, and you start to surf into the physical world entirely. Part of its solution, and what we see customers doing, is they just put a TV on the production floor, they see their machines there, and when an alert comes in, everyone sees it on the TV, in the break room. You start to kind of mix between worlds, and there a very deep understanding of the customer is required. How does he live? What does he do? Which is really true, and how do you do a user interview for a customer like this? So I really love visiting customers, because, as I said earlier, I really love manufacturing, I really love getting to these places and communicating with people, so we visit customers a lot, in parentheses, we take advantage of the customers we have in Israel — and that's quite a few customers — and we fly all the time to Europe and to the United States, and every customer call of the Customer Success teams, I take advantage of it and jump in. In general, customers have a touchpoint, weekly or bi-weekly, with their Success person, who helps them kind of talk, retention,

Eyal David: [16:03] and understand the problems, so it's a very good opportunity for our product people to join. That's it, and as mentioned again, your previous role gave you tons of understanding of the world of those customers you serve, so amazing. So let's really understand for a second who we're talking to, let's dive a bit into your 'why', so Shay, you basically came, as you said, from your degree, and we also talked about your trip at some point. What a trip, what a period. Wow, wow. And you basically entered a world of employment, that in the end led you to Unilever.

Shay Shitrit: [16:33] How did you get to startups, how was it? So I always had the entrepreneurial-tech bug. Already in my last two years at the Technion, I managed the entrepreneurship program of the Technion Alumni Association and the Agassi family. I basically helped Technion-alumni entrepreneurs get leading mentors from the industry, integrated them into the venture's life. There, by the way, for the first time I also heard about Augury. Like, when I joined Augury, I saw in my old emails the first-year summary of Augury. I sent it to one of the founders, it was kind of a mind-trip. How did you even get to this role? So a friend of mine, who did the role before me, is someone who was with me in the army, and sometimes, like in general, I say, throughout my life there are all kinds of opportunities like this, that just pop up and come about, if you just say yes and listen with an open mind, things kind of roll along. And he just said to me, listen, I have something that seems to me to have a future for you, it's such a waste not to. Come, meet for a sec whoever's responsible for this project, for this program, that's basically Ruben Agassi, in partnership with the Agassi family. And come, let's roll from there. And there was a very good click there, and a ten, you know, it's history. And then you basically entered the whole tech aspect from there? Very much, very much. It opened up my way of looking at things, and being kind of in contact with the industry, and conferences, and hearing, and I said, wow, wow. That's why at the end of my degree, where I actually went to the worlds of industry and manufacturing, it was against everything that people kind of thought at that time — 'Shay, you should go to startups,' they actually thought you'd go to startups. To startups. And like, you kept the job, so you stayed

Eyal David: [18:13] kind of in the industrial world, as you called it, for how long? Five years. Five years, quite a lot. And then,

Shay Shitrit: [18:20] you basically moved to a startup, how was it? Yeah, somewhere where it started, the bug kind of, or the impatience, to come and build something myself — over time I discovered that I really love to build. I really love building products, relationships, and even small things at home, I've said I'm pretty bad at that. But I love to do, I love to create. And then, after a few years at Unilever, I felt it starting to bubble up inside me, to the point where I just couldn't resist it, and I just jumped, like really jumped, and together with a partner, we opened a venture, that in the end didn't succeed commercially, but it was an amazing period, and an amazing experience. How long did you run it? A year and a bit. We called it Omnis, which means 'everything', or 'everyone' in Latin, we were influenced at the time, and basically it was in the fintech field, we came with an approach that basically the biggest enemy of a trader in the stock market is basically the trader himself. Because he's driven by emotions, by his own thoughts with himself, and that affects his transactions. So we built an algorithm that monitors thousands of traders in the stock market, knows how to give some score to each of their transactions, loss or profit, and so on, and then just attach money to whoever's at the top of the table. Okay, and then basically,

Eyal David: [19:42] it didn't succeed, as you said — I'll take it from here, and then you basically got the opportunity to enter Augury, how was it?

Shay Shitrit: [19:50] It's another one of those rolling events, Augury, so again, it's something — a company that always kind of, I knew about, and it's well known in Haifa, and it was somewhere there, in the back of my mind. What year are we talking about, just? We're talking about 2020, and... one. Okay, good period. Really, early 2021, exactly. And the guy we rented the apartment from, Noam Green, another very special person, listen, was involved, forms a very good connection — again, at Omnis. At Omnis, when I was at Omnis, yeah, and he was my landlord, we'd sit over a beer, he'd challenge me about all kinds of things, and then I said to him, Noam, I'm closing Omnis, oh, wow, okay, excellent. A week after that I was already at interviews at Augury, in operations, finance and Customer Success, and something in Customer Success, from the very first second, felt the most right for me. The work with customers, the background I come from, the worlds of manufacturing and planning, felt like

Eyal David: [20:56] a real glove. So wait, I'll dwell on this — so he basically brought you into it, without talking about product at all. What was, by the way, his role at Augury, at that time? Group Product Manager. Oh, nice. And then he basically brought you in, but you didn't plan to get to product, or you did, right?

Shay Shitrit: [21:11] No, I didn't plan to. At that time, I really fell in love with Customer Success, and I still love it. It's an amazing and very meaningful function in organizations. Especially at our place. Really. Meaningful, really. I was a Customer Success Manager, and then I got to lead the team, and recruit, and build processes, and it was a period that was amazing. But again, again, that place of building and creating, and touching technology, and multi-disciplinary teams, something... bubbled up, again. Again it bubbled up. Same thing, same cyclicality. And again, Noam kind of, listen, there's an opportunity to move to product, does it interest you? Yes?

Eyal David: [21:59] Okay, let's see. Boom, and the rest is history. What's 'boom'? Was it like a click, or was it kind of a quick back-and-forth? How was it actually? How long did it take?

Shay Shitrit: [22:06] There was a transition period. There was a period where, because of my role in Customer Success, I did both roles in parallel, for a total of two months. But, meanwhile, you know, I read Marty Cagan, and got kind of into things, and building the... so how long does the onboarding of something like this look? Until you become operational? I'd say between two and three months, but even then, you're kind of — you don't really understand. Until you put out the first initiatives, and you don't do these big things, you don't really understand. I look back now, you know, in the mirror of time, and I say how much I grew, and how much I developed in this world, where the discourse now is on another level, but I still feel,

Eyal David: [22:52] that I'm still standing, I'm already standing on it over time. It's cool that you're so modest, and that's great — I'll tell you later too, it's probably like this, and it's a good approach, it's also an approach you believe in at your place, I remember from the marketing stuff, but like it happened here, just — there's another correlation here, of someone again believing in you, right? It also comes back, there's a 'dad' there in all of it, and basically people believe in you, and read good things. Okay, so two years have passed, and now let's actually talk about the subject that we actually gathered in honor of — this is the subject we wanted to talk about, about bringing AI tools basically into the product department's use, for efficiency, something you lead at Augury,

Shay Shitrit: [23:27] you do it with someone else, right? Yes, with Dov, an awesome partner. So I'm already an old groupie of Generative AI, even before GPT came out, I'd use OpenAI's APIs to build all kinds of products, for me and for friends, so it was like totally insane, because people didn't even know it existed, it was really cool. And when I came back, I was drafted for October for four months, and when I came back, I said okay, I was even hungrier to build something more, I also moved squad at that time, and then I said to Dov, come on, let's fly on Generative AI at Augury. We started kind of mapping the central areas that a product manager basically touches, the discovery stage, the spec, the release notes, and so on, and our approach was, okay, so we can do data analysis on things, to surface insights, to surface all kinds of things we'd miss — we'd analyze it ourselves, which is based on text — and we can also create content. We basically chose what the main things are, that'll bring the biggest impact, and will also be kind of the best for an MVP, in the method of iterations, and we decided to touch two things. One, everything related to the worlds of discovery. That is, to look at areas where we want to understand the problem there, and the opportunity, in depth, to use all the information we have, to distill the insights from there. We had a few very good cases, that we learned a ton from; recently there was some really really good use case, where we tried to understand how we can make more efficient all the time and the stage between signing a contract, with a customer's leadership, in a corporate persona,

Eyal David: [25:23] and the actual installation, in the plants themselves. Is that some leading KPI you have, the moment of installation, that's where data starts to flow from, and things happen, right? Yes, it's something very meaningful. Like, in the customer lifecycle, that's the meaning. Yes, very, exactly.

Shay Shitrit: [25:36] Because we have physical hardware that we install, it's very important. And then we said, okay, what information do we have, that can serve, for a sec, the understanding of the problem. So we took a ton of calls with customers around these topics, from Gong. We took from there the whole transcript, of lots and lots of calls. We took content from Slack conversations, context from Slack. We took call summaries, summaries of calls from the CSMs and the RSMs. We took materials, presentations, docs, about the new work processes we're building to improve. And then we said, okay,

Eyal David: [26:14] let's try to distill insights. And it was... so just a modest note again, we did it, five minutes passed, one sec, I have to understand. I know with GPT, that I can upload X things, and after a while, it yells at me, well, too much content, calm down. What does it look like in practice? You took all this goodness, collected it, okay — I can also do now, what?

Shay Shitrit: [26:34] Yes. So basically we used Claude here, whose context is a large context. We're talking about a very very large context, and there the preparation work was, let's call it, Sisyphean. Still, your ability to do this yourself, without Generative AI, is of course nonexistent. So you invest the three, four hours, in all the preparation of the infrastructure, of collecting the information, and putting it into the models, but afterward, what comes out of there — that's it. Okay, come, for comparison's sake,

Eyal David: [27:04] if we wanted to estimate how long this would take today, it would probably take collection work, after that basically someone who'd go over it a first time, fails to produce from it, you try in a bigger group, nothing comes of it, you do an off-site or two, nothing comes of it, or it does, whereas you compressed it into just a few hours. So from the moment you basically got output now, something that makes sense,

Shay Shitrit: [27:29] some more massaging — what does it look like from there? So yeah, the next stage was basically to see that we're kind of not way off. We started talking kind of internally, internal stakeholders, and the agreement was so broad, that it gave us the best stamp we could get,

Eyal David: [27:44] that we're in the right direction. So did it discover something, or did it basically point to something everyone already knew? How was it? Or did it connect together things everyone already knew?

Shay Shitrit: [27:52] It did both — like it brought new things, new insights that we say, okay, that's very interesting — and because everything was also based on things that were said in a different way, by customers, it could be that it was in the data, but no one said, ah, this is the reason, this is the problem, okay? And then afterward we said, okay, let's see whether at this stage it connects kind of to initiatives that today we already have, okay? Today we cover these areas, some yes, some no, we saw how it connects kind of to things, in the future, and that's currently some very meaningful backlog, that from there we'll start bringing in, to handle this thing, in a process-based way, and in the product way. Wow, well done, so basically it's a kind of,

Eyal David: [28:38] it'll integrate into your roadmap, at the end of the day, that's what you're saying. Unequivocally, unequivocally, and I think that's also the greatness of this tool — that you look at the world of discovery, the ability to surface problems, really give them kind of a chunk — why the un-surfaced problems, even quotes from customers, how many times it was said, who the customer is, and then you connect it, also what the size of the customer is, how meaningful it is, and help you see these blind spots. It's amazing, you can also put in — really not only quotes — you can even get the seconds, the timestamp, from the transcript, which, for that matter, and then if you want to present it, and there's also the matter of cultures here, after all you have customers from across the cultural spectrum, so suddenly it bridges that, looks at net data, it's really interesting. And okay, at the discovery stage, it makes sense, it's something you'll probably do once in a while, right? Did you set when the next time will be, or are you currently swamped with the roadmap? I'm probably not going to do any discovery, without using this tool. Yeah, so we'll also talk in a bit, I think — when you set out, so you really defined a few things you want to improve on, one of the things, or that they take into account, when you basically do this process, you didn't talk about adoption — that's maybe something worth understanding for a sec, what it looks like inside the organization, inside the group more precisely, but let's tell, so okay, you talked about discovery, and you basically identified more aspects, right? You talked about specs, about release notes,

Shay Shitrit: [30:02] something else? So the thing we really focused on is really the spec generator. We built, in our Azure dev environment, an agent that basically has very very broad context, about Augury, about the structure of the fleets, the squads,

Eyal David: [30:17] what the role and metrics of each... what do you mean we built? Who built it? Me and Dov. How was it? What did you do?

Shay Shitrit: [30:25] So we really defined it, we said, okay, what do we want to output at the edge? We want to output a spec that knows how to bring a plan, an action plan, and to define the problem, to define the solution, and what we basically want to measure here. So what information does it need for that? What information needs to be fed into the model? So this information, we defined it in the form of: what's the specialty of each fleet, of each squad, what they want to solve, what the strategic plans of Augury are, so that it'll have very very broad context, initiative — those are specs, for example. Probably the resources at my disposal, right? Right, exactly, really. And then this agent became not only 'tell me and I'll output a spec for you', but a brainstorming process, that asks you for information along the way, it knows how to fill in what's missing, and you tell it, fill this in for me for a sec, fill this in for me for a sec, and in the end to output some spec for you, that tells a very nice story, tailored to your domain. Which again is amazing,

Eyal David: [31:23] but again I ask, if you reinforce this thing, what does it look like day-to-day, like the product department — tell us a bit what it looks like.

Shay Shitrit: [31:31] So the moment we built this tool, we presented it for a sec to all the product teams, we said like what information it needs, what it knows, what it knows how to offer in the end, and we just gave free use. So I'll say that for the spec generator, we see greater usage than for the discovery areas. The discovery areas, the discovery areas, the discovery areas just still have a not-small friction point, and that meets us in general in the adoption of technologies. So it's not only to change my work process, there's just also friction for me, in order to build the discovery I need to invest three, four hours, and well, I don't know in the end what I'll get out of it,

Eyal David: [32:13] what the value of it is. I wasn't convinced. So should I invest this time now. That's it, I'm kind of thinking out loud — wow, maybe it's worth it for Augury to hire someone internally, whose job is that, who's kind of the gatekeeper of the Gen AI, basically like there used to be a librarian, for that matter, who goes, comes back with the data, and interesting, maybe they'll hear us and all. And I think this is also related, really like you said, like there are processes that happen basically all the time, like spec creation, where the ongoing part is, and there's basically — but how do you measure it?

Shay Shitrit: [32:44] Is it like from the feedback you hear from the group? Feedback — you can also see how many chats were kind of opened, and by whom they were opened, but mainly feedback of, it helped me here, I produced — here, I produced the spec here,

Eyal David: [32:57] people share it a lot. I think also with specs, after all it probably also helps you foresee basically all the risks, or most of the risks that can surface, which again, as a human, is harder, right? And is this thing connected to more bodies in the company, like, that have their own agents? What does it look like? So we built another agent,

Shay Shitrit: [33:18] for general — did you build that? That's more Dov, that's more Dov, Dov said, yeah, me and Dov, power couple. Ask Augury, where basically there's more context, which is also marketing context, and this marketing context, we put less into the spec, so it wouldn't create kind of some irrelevant things, that we're going to do in the future and don't really exist yet — but for the Customer Success teams,

Eyal David: [33:46] or anyone who wants to ask about Augury, whatever they want. Amazing. And watch, what does such a process look like, of kicking off within an organization, because you basically described a situation where there was no one who did this, you came in with Dov, tick tick tick, you kind of ticked away, some PoC, let's call it fast — I don't know how fast it was — and you basically started presenting results, which is amazing. What do you do from this moment basically,

Shay Shitrit: [34:06] how do you bring this inside? So in the first stage, it's spread the word. It's kind of, start to say, it exists, guys, this is what it's for, see the output, and share the success stories. So the story of the good discovery was such a success story, that circulated in the company. Wow, how did we manage to produce, in a relatively short time, such good insights. And when you do the spec, then you say, we did this with the help of the spec generator, inside the PM — our PM internal channels, or the product guild. And it gives motivation, because it kind of lowers the risk, and other people's perception of the value. They say, wow, this is amazing, how long did it take to make this spec? What, an hour? An hour? Whoa, okay, well, I'll try, I'll give it time, what do I care. And then the moment you kind of also generate a kind of FOMO feeling, because come on, see what a product came out here on the other side, see, these are insights that changed the whole roadmap, and you also say, well, it's not such an investment, it's not such a big risk, let's try it. And you make it accessible to everyone — what does it mean, make it accessible to everyone? Really, a link, it's vibe, totally vibey, enter this link, start talking with this agent, and see what comes out for you on the other side.

Eyal David: [35:19] I understand, you didn't stand in support of this thing in any way, basically, cool. And nonetheless, like, they used it. Right. Cool. If you hadn't done this, to start it off, would it have caught on? You need to help at the beginning,

Shay Shitrit: [35:33] to adopt kind of different work processes — on its own it doesn't happen. Surely there's some entry cost, one way or another, or you're not sure what value you'll get on the other side, so it wouldn't have succeeded. Yeah. And now you want to encourage basically more bodies in the company to use Gen AI, what does it look like? Yes, we're starting — the Gen AI, there's a lot of talk about it, kind of at Augury, there's a lot of talk about how the internal tools participated, how teams used Gen AI to make work processes more efficient, and also inside the product. And it starts to open everyone's mind. So when we sit kind of and talk Gen AI, Gen AI, then maybe actually — things that maybe we couldn't do before Gen AI — it just opens up some thing in your mind where you define that everything is possible. And then when you start doing brainstorming conversations, the ideas that come, ideas like, okay, wow, yes, let's do it, all kinds of crazy things like this, and then you say, okay, that's the edge, but let's do an MVP,

Eyal David: [36:36] to start. But it's something in people's thinking, in the mindset, that opened up. I think also at your place, it's also something I remember, because the focus was all the time on — or still is — on the problem itself, right? You don't talk solution. Yeah, which is great. So basically you use it a lot to sharpen and understand the problem — is that mainly the thing? Very much. And does it have enough context? We add a ton to it.

Shay Shitrit: [37:00] So for the spec it has a ton of context, and you can add to it along the conversation whatever you want, and in areas kind of of discovery, we bombard it with all the context we can give it. Wow. It's amazing. And at your place, more also in development they make use

Eyal David: [37:14] and that's a long time already, so like, amazing — of the product, it takes so much time, really. Cool. Something else you want to add about Gen AI, or shall we move on from here? Because you talked about release notes, but I think it's pretty clear in this aspect. Yeah. So is this something you now kind of measure at the company level, at the department level? What does it look like? So it's not measured,

Shay Shitrit: [37:36] for now we're kind of, guys, here are the tools, use them, and the eyes are now set basically on its implementation in the platform, which Unlocks New Value kind of on another level, and also provides us with a ton of insights about what interests our customers to know. And... naturally, all this information is a kind of circle, it'll enter back into the Gen AI, it'll come back

Eyal David: [38:00] with value anew — no, I'm kidding — but it's kind of a circle, amazing. So, well, we're already converging toward the end of the episode, I'll ask you, Shay, where do you see yourself, what do you wish for yourself in ten years, where do you see yourself? Oh, excellent question.

Shay Shitrit: [38:14] You know, my wife, Yulia, and friends and family, will attest that I have some sentence that I always say — I leave it to future Shay. A thing I say, half say, half not. By the way, future Shay is better, he's smarter, he has more experience, he has a different perspective. And that's what I wish for myself — it's just that future Shay will always keep improving and being better, better to people, better to himself, smarter, go with his heart, with the courage, just keep going as he is. Come, come let me tell you what'll happen, there'll be bubbling, and then let's see.

Eyal David: [38:55] Cool, and I wish you — what happened is what'll be. And if you have some piece of, a gesture — recently — some piece of advice for someone who wants, or a woman who wants to get into product, and she's either already working in traditional industry, or she's in general, or he, works inside a tech company specifically, and they want to kind of get into a product role. How do you recommend doing it?

Shay Shitrit: [39:17] I think people is the key to everything. Create it from an authentic place, from a place very real and deep, these things, and put your desires out there, expose yourselves. Don't be afraid, go and be brave. And in the end, things work out, in the right directions, if you're open, if you kind of help things roll along, to roll into the right place, in the end it works out. That's it, it must be said, you were again

Eyal David: [39:45] modest, basically in your transition to product — it's not an easy transition. There's a ton of material to learn, there are a ton of conversations to have, it's a very hard onboarding, especially if you come without a background, and you basically also mentioned that Yulia actually worked, your wife worked, right, she didn't sit at home, she worked full-time, and she's very busy, and you're also busy in the previous role, and you have kids at home, and you kind of busted your ass, and basically worked at this to make it happen. So like, that's what you're saying — rest assured, you don't mind that I'm interpreting.

Shay Shitrit: [40:15] Sure, sure, it's a ton of hard work, no doubt, and when it falls into the place of your true and deep passion, you bring out the energies and the abilities and the way to put it into action, right, but the very deep place of emotional connections — it starts from there, but

Eyal David: [40:35] it's not the only thing that matters. So you took it, and put it into action, with all the difficulty and challenge in it. Love it. Great, Shay, thank you so much, it was a pleasure. Thank you so much, me too, really. Come on friends, until next time. Bye bye. Hi friends, thanks for listening. If you found this podcast valuable, you can subscribe, follow us, and of course for more episodes, on Spotify, Apple Podcasts, or any other app. Of course if you didn't find us on some app, I'd be glad if you'd write to us. We'd love five stars on every platform, and for you to follow us, so that more listeners can be exposed to us, and find the podcast. You can also find the previous episodes, on any app, or on the YouTube channel — we have links in the description. Until — until next time, come on, be efficient, and bye bye.