Why Great AI Solutions Start with Listening: Navigating the Data Quality Crisis with Joe Reis Part 2

Speaker A: Hello, we are bringing Data and AI to Life, a podcast by Informatica. I'm Amy Horowitz, our VP for solution sales for data integration and data governance.

Speaker B: And I'm Nick Dobbins, worldwide VP and uh, field CTO here at Informatica.

Speaker A: If you've ever felt lost in the chaos of data and AI, you've come to the right place. We'll be conversing with industry experts who are here to shed light on the challenges all rocked up within these arenas. We're here to bring clarity to the chaos myth, busting the confusing parts and providing insights and guidance complex data problems by delivering trusted data for analytics and AI.

Speaker B: All right, Lesters, welcome back. In the first part of our conversation, best selling author of the Fundamentals of Data Engineering, Joe Reese took us on a journey from his roots in data science to his deep focus on data engineering. We discussed a lot, but he really cut through that industry hype and he separated the latest trends from the buzzwords and, and redefined a modern data engineer's role. So, you know, we closed that introduction to powerful new concepts like action oriented architecture and the new postmodern data stack. Now let's chat with Joe and dive into how these ideas actually come to life. How do we actually apply this? And kind of tricks me back to like I read through the substack that you have. You emphasize a problem first approach. I think this kind of gets into some of that business value, right? Like instead of just getting tools and go tool first to just see what cool things can you do, what problem do you want to actually solve? And for the data engineer that uh, are out there or the data teams, if we want to go data practitioner, even larger, what is some practical advice to be better at getting to that problem first? Translating the business problem and going out and building that AI solution.

Speaker C: The cheapest thing to do is use your ears and uh, listen to people, right? If you're able to do that or pay attention to what they're saying in writing. If you can't hear for some reason, but it's like these are the things. Just pay attention to what people want and also sort of use second, third order thinking to understand the question behind the question. Like if they're saying they need this, why? Right? Like root cause analysis and I think attacking those things and understanding in your organization that you work in what is the true north of your company. All too often we're focused on sort of the immediate needs of today and uh, depending on where you work, that's totally fine. But if you take a step back and understand the context in which you're delivering something and sort of the bigger picture and whether or not it actually does satisfy people's needs, that I think goes a long way. I'm recording this as I listen to a construction crew across the street. It's actually shaking my house right now as they, uh, are flattening the ground. And that seems like it would be a really fun machine to work. Actually. I would love to go do that all day. If I showed up to your house right now and started doing that, would I be solving any problem for you in particular? Probably not. It would seem really fun.

Speaker B: We might be creating a problem for me.

Speaker C: Exactly right. Like, I think, are they going to like, shake? Is my house going to fall apart soon with the harmonic frequencies and stuff? So it's like these are things where it's just, you got to take a step back and have empathy for people, I think, and understand what their problems are. I think all too often with technologists, we're looking at it from ourselves out where we're thinking, okay, so we need to use technologies. This technology would be awesome. To have my resume, we have to absolutely include whatever project we're doing, whether or not it makes any sense. And so these were sort of the things in the back of your head. I've been guilty of this too. Obviously, you know, taking a tool driven approach first, resume driven development first, it's a lot of fun. But at the end of the day, you might have satisfied somebody's needs or you might just be maybe irritated somebody. Like if I showed up again with a bunch of construction equipment in your front yard and started going to work for no apparent reason, like, it looks like you need this. And it's like, no, actually I don't need this at all. But this is what happens in a lot of companies a lot of times. Yeah.

Speaker B: Well, I could say I just had my irrigation system repaired and I'm glad they didn't show up with an excavator, even though that would be really cool to use. I'm glad they just brought the shovel.

Speaker C: That would have been overkill.

Speaker B: Yeah. As we go with it. But yeah, I mean, you look back, like we actually just had an internal team discussion around. One of the best things that we could do to make things more effective is a little bit of active listening up front. And it's shocking.

Speaker C: Right.

Speaker B: Like we carry our ears around with us all the time. Or for those that struggle, devices that are out there that can help you.

Speaker C: Well, the cool thing Is now. You know, I was talking to some data modelers about this and data modeling is one of these practices is notoriously slow. Nobody wants to be in these meetings, right? I mean it's just kind of the reality of it. And like, look, you have the ability to record conversations and use LLMs to synthesize things and summarize and come up with action points, obviously review it, sometimes it makes stuff up, but like this can speed up the work. But there's no excuse for anybody not actively listening and participating in the conversation these days, especially when you have technology. As sort of the extra assistant to all this, it's interesting. I asked a really wealthy friend of mine once, he was my old boss, asked him one day, how do you make a lot of money? And he's like, well, you sell people things that they want to buy. And I was like, yeah, I guess you're right. Actually. Then he also followed up and he said, you know what? And when you're taking the sale, shut up, don't say anything, listen. But all too often we want to keep adding more stuff on. Like maybe do you want this or do you want that? And it's like, no, they asked for this. That's it.

Speaker B: It's great advice. And going back to a few of the other comments right. You made, we'll go back to fundamentals. Data governance is a big topic for us, but I think it's probably an over generalization. But a lot of times when we are interacting with data engineers, right, like it's kind of a boring, dry, cumbersome type topic. I want to get to my end result and we'll figure out that stuff later. Do you have any tips now that you've been in it? And I think you kind of see the value in prioritizing some of these fundamentals and how do we communicate that better? How do we build that mindset, that lack, uh, of governance, lack of some of these fundamentals actually can hinder innovation in the world of AI in the end versus accelerating big time.

Speaker C: It's funny because you want to avoid the boring stuff except when it counts, except when it's going to help you do the really cool stuff, right? It's sort of like, you know, I have to tell my kids this, they're both athletes. And I'm like, yeah, uh, you know, you have to actually go do the workouts, the boring ones like agility stuff and whatever conditioning. It's like you don't like it. That's too bad because to make you a good athlete, you're gonna have to do this. And it's no different any other endeavor. It's the sets and the reps of the boring stuff that get you to the point where you can do the cool stuff or you can get the touchdown or the, the big pass at the end of the game and it's like, or win the wrestling match, whatever, but it's the boring stuff, in and out that makes the wins. But how do you communicate this to the business? Right? Because nobody wants to have that conversation because that's super boring. So here's a couple tips. I found that data projects win when you don't talk about data and when you use terms in the other person's language because then they feel like, okay, so we're doing a project, we're not doing a quote, data project, which sounds super boring and super nerdy. Nobody wants to do that. You do, because you're a data person. Not unique, but maybe you do. But like the general audience here, we are used to nerding out on the stuff we're interested in, right? Which is all the data stuff, all the technical stuff, all the stuff that nobody else in the world really cares about. Example, right? If you were to go down the street and just tell a random person what you do for a living or the industry you're in, they're like, that's a made up thing. That's not even a real job. But you know, if you mention things like marketing or sales to people, they're like, oh, okay, you advertise and you make me buy stuff, great. But these are the sorts of things where I think if you can talk in the language of somebody else and help them understand the types of problems they're facing and help them solve, but try not to use the word data at all actually, and try and couch the problem and the solution in terms that they can understand. I've seen that this is actually a winning solution in working with basically anybody on a project. And so that's something that I was always trying to do was just. My old data engineering practice is just strangely enough, don't try not to talk about data. Maybe have your punch card of like the times you can say it or something, but just like keep that to a minimum because nobody really cares except you.

Speaker B: Yeah, it's interesting, I mean, because literally, you know, we go back to, I just mentioned, right? When you're talking about listening and active listening was one of our big emphasizing. And the second part of it is mirroring back. So essentially, as you hear what's important to Them and the language that they use. You know what you're talking like data people know data, others don't. So it's your job to translate back and mirror back what you are going to provide in their world to the value. So I think it's great advice as we go with it.

Speaker C: But it's hard though, right? This is where I do urge a lot of data people to become business literate, understand business terms, business concepts. There's no disadvantage to it. My friend Jordan Morrow has a new book out. It's about how data people can speak business. There's also really good books just with the visual MBA and 10 day MBA and stuff like invest your time into understanding these concepts. Plus I think it just makes you more effective in like, in general, in life, just understanding how business works. But until you have that empathy, if you're talking to an accountant, for example, and you know you're not using their terms or like, okay, now they're having to educate you and now they're having to feel like they're doing a lot of work. Definitely.

Speaker B: Well, hey, as we end time, there's normally two kind of things that I always like to look at as we go with and kind of pick your brain. So one, it's kind of the look ahead. So lots of excitement in the data IIS space. What's one piece of advice for those that are currently in the field to really thrive, right? Like not just survive, but really get after it.

Speaker C: It's an interesting question because I feel like we're at as boring as I sound like. I feel like this is probably one of the most exciting periods that I've seen in decades. And I wrote about this maybe, uh, a few weeks ago where I felt like a reset button has been hit on the entire industry, not just our industry, but like a lot of industries actually. And we're all trying to figure out what this looks like. Like we're all collectively staring off into the void, wondering what's on the other side? Is there AGI on the other side or some super intelligence? They'll just do everything for us? I have no idea. Nobody does. But what I would say is master the fundamentals, right? That uh, I think this is going to provide good grounding. But apart from that, master the new ways of doing things and pave the way for yourself. Like because a reset button has been hit, you have the opportunity to invent new futures in the industry that don't exist yet. It reminds me a lot of what happened with the web, where all of a sudden There were webmasters and all these other rules that just didn't exist before. This is that opportunity right now. And so I would say don't look at it with a sense of trepidation like, oh my gosh, my job is you can invent a new job now. You can invent new ways of doing stuff, you can invent new infrastructures. And this is the message I'm giving to people, is don't passively sit here waiting for the future to happen to you. You have the ability to create a new future right now and change the industry. And this is an awesome time. So I'm like just beyond excited about the possibilities of this. I think it's rare.

Speaker B: Yeah, I agree. I mean, it's an expiring time as long as you're. As long as, yeah, you have the fundamentals, you have the skills, and you have the willingness to embrace the change. Go make it happen. So the other one, right, like you've seen a lot, you made a transition as a, uh, recovering data scientist to a data engineer as well. We also get a lot of listeners that are looking to get into this space, whether they're, you know, in college and looking to see how do we break into the things, or whether it's, you know, retooling and finding how should people get started? What's the best way to kind of get going, getting into this.

Speaker C: It's an interesting question right now because I feel like I advise a lot of universities and I talked to a lot of professors. I was just talking to one the other day. And the schools are equally confused right now, which is funny because they're trying to figure out the existential raison d' entre of a university these days. So everyone's trying to figure this out, but I would say a lot of the education is going to have to happen on your own. If I were to take a step back and understand what I think is necessary these days, actually, philosophy and the arts, I think are going to be more important than ever. In addition to understanding engineering, because you're dealing with something that is language based. Now this is the big difference where I think the intersections of the fields are interesting, where now data is crossing into knowledge and library sciences, I would say this is the next frontier of where as a practitioner be looking from getting to the field is the data stuff is great, it's great and foundational, but there's going to be a lot more because you're interacting with machines that can synthesize, summarize and speak and quote, reason in their own way. But this requires a fundamentally different toolkit that up to now we haven't been taught in computer science or data. And so I would say the knowledge sciences, information, uh, and so forth is where I would also look at. I think philosophy provides a really good foundation for all that as well.

Speaker B: It's an interesting take. I haven't heard that one before, but it makes a lot of sense. Right? Like emotions and reactions are all part of the world now.

Speaker C: It is. And what's interesting now is graphs like knowledge graphs and semantics, all these ontologies and taxonomies, these are all the words that are entering the zeitgeist in the data world. This wasn't the case a couple years ago, but everybody now has seems to have a semantic layer, everyone seems to have a graph. Interesting thought experiment I've been telling people to do is actually ask AI itself. Ask your favorite chatbot if you were to design an architecture and a data model that would fit for an AI first world, what would that look like? And that is freaky, actually. You ask all of them, they come up with basically the same architecture. It's a graph. It's a real time graph heavy in semantics and a bunch of these other things that we just don't really, we haven't really paid attention to now or this is all becoming front and centers. So the intersections are what's most fascinating. But if you're getting into the field, I would say enter it with an open mind. It's probably going to require a lot of self study. There's a lot of things I would say for courses, if you want, obviously take my course in Coursera. But deep learning AI Andrew Ng's doing a terrific job with a lot of the stuff he's doing. There's a ton of short courses he's doing every week and they're amazing. And that's at least to get you up to speed on a lot of the new tools. Because as we talked about at the beginning, the amount of tools coming out is absolutely dizzying. Like I don't think you or I can pick. I don't, I can't keep up.

Speaker B: So no, I'm not even going to try.

Speaker C: It's so hard. But that's the advice I would say. But you got to definitely step outside the comfort zone these days. I would say, ironically, in an age where everything is provided by LLMs, read a lot, read wide and read widely. I think that building the mental models for yourself is going to be a good investment in terms of navigating this new world where, um, AI is going to be increasingly embedded in pretty much every fabric of your existence. But how do you know what's true? This goes back to the philosophy part, right? It's why I say that you would have to understand that, because if you don't understand what's true, then how would you know anything?

Speaker B: No, that's true. I mean, I tell a lot of the different college kids that I talk to on this and that I'm like, well, you know, as you go out, I think the two biggest things you get right is build a core. So I guess kind of, you know, learn your fundamentals. And I said, and be very comfortable. Embracing change is the. The pace and cycles of change are just accelerating like crazy. So, you know, embrace it and be comfortable with it. Take advantage of it. Then you'll survive. You'll do well.

Speaker C: Do well or learn plumbing. One of the two. Yeah, that's.

Speaker B: Yeah. Or be a plumber. I mean, you will always need that. But I feel like we could probably put here and banter on and talk about this for the rest of the day, but time has come. Joe, thanks for. It was a great conversation. Tons of good insight to share out there. So hope the listeners loved it. I can't wait to talk to you on the next one.

Speaker C: Thanks very much. Thanks, Nick.

Speaker B: Bringing data and A.I. uh, to life is brought to you by Informatica. To find out more about Informatica and how our intelligent data management cloud can help you achieve better business outcomes, head ah to informatica.com.

Speaker A: stay tuned for more illuminating discussions until we meet. Next time, keep harnessing the power of data and AI to bring transformative outcomes to your life and business. Make sure to click subscribe so you don't miss any future episodes. And tell your friends about us too. On behalf of the team at Informatica, thank you so much for listening.

Speaker C: Sam.