Andy Vitale:
I think the worst thing that you can do is tell people to try to like avoid it or lean into like the wrong areas because it's here and companies are adopting it, they're embracing it, they're pushing for it. And if you're one of the people that's seen as, you know, more of a speed bump than an accelerant, that's gonna create a problem for you.
Kristiana Corona:
Ever feel like everyone else has leadership figured out and you're just making it up as you go? I've been there. I spent two decades leading design and technology teams at Fortune 500 companies, and for years, I looked like I had everything pulled together on the outside. But on the inside, I felt burned out, overwhelmed, and unworthy of the title leader.
Then, a surprise encounter with executive coaching changed my life and dramatically improved my leadership style and my results. Now, I help others make that same shift in their leadership. This podcast is where we do the work, building the mindset, the coaching skills, and the confidence to lead with clarity and authenticity, and to finally feel worthy to lead from the inside out
Hi, and welcome back to the Worthy to Lead podcast. I'm your host, Christiana Corona. If you're a design, product, or engineering leader, I have a question for you. How much time are you spending actively learning how to code in AI tools each week? I'm not talking about taking your copy and putting it into Claude and having it write your emails for you I'm talking about rolling up your sleeves, getting uncomfortable, and learning something new time.
A couple hours? Or maybe none at all. Maybe you're quietly hoping that your team is gonna do the work, and they're gonna bring the best practices back to you, and then you will know how to lead in this new world. Well, today's guest does something that might shock you a little. He spends 75% of his week actively engaging in and learning about AI tools at the VP level.
I'm joined today by Andy Vitale, VP and Head of Product Design and AI Center for Excellence at TaxWell, and one of the most respected voices in design leadership today. Andy is a seasoned design executive whose career spans healthcare, financial services, and enterprise technology, with leadership roles at Constant Contact, Rocket Companies, Truist, and 3M, where he was already designing AI-powered systems nearly a decade ago.
He's also a founding member of the Design Executive Council and an adjunct professor at Kent State University. He's a globally recognized speaker who has appeared at events including South by Southwest, Adobe Max, SAP Design Talks, and more than 100 conferences worldwide, speaking on leadership and responsible design, AI transformation, and the future of work.
In this conversation, we're gonna trace the arc all the way back to 2015, where he was doing auto-suggesting medical codes at 3M, to then billion-dollar conversational AI wins at Rocket, to now shipping TaxWell's innovative new tools in a single tax season. We talk about how product teams are working differently, how vital design systems are, and why leaders who treat AI as something to delegate are quickly becoming speed bumps instead of accelerators.
Get ready for a fun and always interesting conversation with Andy Vitale. Let's dive in. Andy, thank you so much for taking time to be with us on the Worthy Tuli Podcast. We're so glad to have you here.
Andy Vitale:
Of course. Thanks so much for having me on.
Kristiana Corona:
So, um, you and I were catching up, uh, a little while ago and kinda reminiscing about some of the past work that you've done, and you and I had met, I think, several years ago when you were still at Rocket.
But a lot of the work that you've been doing around AI recently, it was interesting 'cause you were like, "I've been working with AI forever." Like, for I think since 2015, right? Back when you were at 3M. Yeah. Yeah,
Andy Vitale:
Exactly.
Kristiana Corona:
And so this isn't anything new or scary or crazy for you. You have sort of this track record of working with it over time.
So I was curious if you could maybe walk us through what is your history with AI? Like, how have you been using it since 2015, and what was that like before it kinda got this big, uh, boost of notoriety?
Andy Vitale:
Yeah. So I, like you said, I've been using AI as a material since, uh, 2015 when I was working in healthcare, and we were using artificial intelligence, machine learning, natural language processing originally to auto-suggest medical codes for people that do code documentation, medical coding to effectively and efficiently get bills out the door at hospitals.
And we started for outpatient facilities and showed a really interesting increase in productivity, and then we started to focus on inpatient software. So this was a time where we were really looking at People that were doing this for a living that had memorized these diagnoses codes, and they would enter them every time, you know, they were coding a chart.
And what was interesting is when we started to introduce this auto-suggested coding, they were fearful, right? Like, is the machine at some point going to take my job? And they didn't want to adopt it at first. But we got lucky because the government changed the level of specification needed and changed the clinical diagnosis number.
So it went from ICD-8 to ICD-9. So when they changed the level of specificity, it went from like the people that used to memorize this documentation had to look it up. So they thought, "Well, maybe I could trust but verify the system and, and see how that works." So we ended up working with AI there and then eventually predictively modeling outcomes in hospitals that measure excess expenditures or improper quality of care.
So that was an interesting foray into AI early, early days to really see the impact. And what's interesting about that is the feeling that our customers had in that moment is the same feeling that our industry is having right now in this moment. Fast-forward further along in my career, I worked in banking, got to really spend a lot of time with like Infosys, and they were doing interesting things around sentiment analysis and how you could capture like a phone conversation with a banker and see if based on the sentiment of someone, how do you change the way you would talk to them?
We ended up i-implementing a lot of interesting tools there around AI. Then I'll take you to Rocket, like you described. At Rocket, one of the opportunities that we really had was on the conversational side. So back then, there wasn't somebody on the line twenty four/seven to answer your call, and we had a system that you would use a-after hours.
And our goal was: how do we make that system smarter to capture people who are ready to, to, you know, start on a mortgage application but needed to get some sort of information to make that happen? So we introduced, uh, IVR, IVA technology, basically routing people to the right place and increased loan volume by billions of dollars, which is, you know, was amazing at the time.
And just seeing how AI started to go from a thing that really not a lot of people knew about unless you were working in technology to it was starting to become: how do we make our product smarter? Then I ended up at Constant Contact, a digital marketing company, when generative AI was really like all of the rage coming out.
How do we bring this into our products? How do we help with our writing? And we saw an opportunity of like people are looking at creating Marketing campaigns, you know, it starts potentially as an email. But once we start to help them draft that email, what's preventing us from turning that email into a full campaign with like a calendar, an SMS campaigns, and social campaigns, and can we help them generate those things?
And now I, I'm at TaxWell, and we handle taxes for millions of tax returns for professionals and filers, and we've got AI in our product. We launched a, a really interesting thing that we call Smart File, where you're able to get help with where you may be stuck in the process. And then we also introduced a connector with Claude, a custom connector, to try to capture people where they are, because we saw that some people this year were using taxes with Perplexity, with Claude, right?
So how do we stay involved in, you know, meeting our customers where they happen to be? And we've also introduced AI into the way that we work, not just for, you know, design product engineering, but across the whole company. We're looking at opportunities to just make ourselves work smarter and also make our products more intelligent.
So for us, it's not abnormal to have designers working in, in Cursor, not just Figma Make, and, you know, have product people also building prototypes like that. So it's really, AI has become the material now in which we work.
Kristiana Corona:
Yeah. You brought up a lot of interesting topics I wanna dive into there, but maybe as a nutshell, like if you think back to some of these ways that you've used AI, there was always this theme of probably some pretty bad customer pain that they were putting in some pretty awful repetitive tasks that you were able to remove from the customer experience or from those agents, um, that were dealing with those things that really moved the needle, and this was, like you said, this was just the clay you were building with or the, the, the tool that you were using at the time.
When you think about just some of those previous experiences, what were your takeaways around like AI being an important part of that, or was it just another tool you were using just like any other tool?
Andy Vitale:
I think it felt at the time like, at least in the beginning, like it was just another tool. It was a way to help people solve problems in a different way that was new to us, just like other emerging technologies that I've worked with over the years, like AR, VR, right?
Even the, the advent of mobile. And For us, it's always, especially with, you know, design or building products, it's how do we seek out that customer problem and what materials do we have, what technologies do we have at our disposal to help them achieve what they're trying to achieve in a better way, whether that be faster or with less errors or in a way that helps them trust the system just a little bit more.
And I think right now, just the advancements, the speed at which we're seeing AI just kind of like Compound on top of itself of learnings and capabilities is something we've never seen before, but we're not building AI or bolting AI on just for the sake of like, "Hey, we've got an AI feature." It's got to really do something that's important for our customers.
And for us today, sometimes that's, "I've got a lot of tax documents. What do I do with all of these?" So how can I start to like ingest that and organize that for people so that it doesn't feel like it's just a tedious task that someone has to do, but it's actually helping them achieve their goal in a faster way.
Kristiana Corona:
That one really resonates with me, uh, being a small business owner now and having lots and lots of facets of documents that I need to, to capture and put in one place. And so as you were describing that, like I think you talked about it on LinkedIn a little bit, your experience, like that was only a 12-month turnaround for your team to say, "Hey, we have some ideas.
We now have AI. You are fairly new at the organization. How do we make something by the tax date next year?" Like how did you do that within one year?
Andy Vitale:
I mean, it's funny because some of the things we talk about happened a lot faster than one year. Like our smart file came together in, in a couple of months, and I think it's just we're, we're trying to always set out ways to solve problems, and we understand like document ingestion is really hard.
So how do we first make something simple like document import? Make it easy to at least find your documents, have a checklist, be able to drop them into one place. And then as we start to see what's available to us, we're like, "Wait." Instead of like manually trying to map these, we can leverage AI to, to help us organize these, and we understand that our customers trust us.
So if it's something that we don't have confidence in its ability to map, then we're not going to put that out there. We're gonna make sure that it's right. So, you know, you start sometimes with a couple of documents, and then you realize, "Wait, with this technology, like we can do multiple types of documents."
So everything starts with an MVP, and we're always trying to find ways to get a product to market pretty quickly, even when it's not tax season, like we are still building. So people won't get to see all of the things that we built until next year, but we're steadily like building and releasing along the way.
So we're always trying to make our product better, and we just have tools now at our disposal that make us do that in a, a much faster way.
Kristiana Corona:
So you're in a very highly regulated industry And that's got to have people concerned about integrating AI at different places and its reliability, and what if it fails, and what if it makes decisions that we didn't agree on, and all of these things related to taxes.
I mean, that's a high-risk type of thing. So how did you have those conversations and help set people at ease that the way you were integrating AI was not going to disrupt the reliability of that product?
Andy Vitale:
Yeah. I mean, our reliability is so important to us. We have all sorts of guarantees for our customers.
So as we think about The capabilities, the conversation happens in, like, day one when we're thinking about these things. We've got legal, we've got compliance, we've got information security all, like, part of this. And, and every step of the way, we've got checks and balances and the right guidance and the right governance in place to make sure.
And we're like, we're not just double, triple-checking. We are checking on a regular basis to make sure that everything that we're doing meets our own standard of, is this safe? Is this secure? Do we trust it? Before we'd even think about putting it in front of our customers and have our customers get to experience it.
So we make sure that we're checking things on a continuous basis. The good thing about taxes and the bad thing about taxes is when we're not in tax season, people don't wanna think about taxes, but everyone that works in the company has to file taxes. So we do have a lot of access to our own users or to people that use our product.
Now, of course, we're a little bit more advanced than others in certain ways, but we have a lot of people that we can continue to test with as well.
Kristiana Corona:
You brought up something really interesting, and I, I've had similar experiences as well, where, you know, maybe product teams are thinking about design, engineering, and product owners coming together and ideating.
Like, that's sort of the, the magic triad, and everyone is always inviting those people. But you mentioned several other roles when you started ideating, when you started concepting, whether that be legal or compliance or tax folks or different people who are often left till the very end as the gatekeeper.
And I feel like that is a brilliant move and something that all of us should be doing more is, who are all the right voices that we need to have from day one? Um, accessibility is another one. So what, what's your philosophy on that?
Andy Vitale:
My philosophy is it's way easier to involve people earlier so that everyone feels like they're part of the solution and has that accountability in the problem that we're solving.
But I also know what it's like to feel like you're brought in last minute, and anyone that's brought in later in the game, they've gotta get caught up. They've gotta understand where we're at. They don't wanna be a speed bump, but at times, like if you didn't involve them early enough, you've gotta get them caught up to speed.
So how do we bring them in early enough so that they move things forward as opposed to-- 'cause no one wants to come in and be like, "Hey, wait, stop," right? "I've gotta get caught up. I'm not on board with this." While I would love to involve 15 different departments on day one, I think that that becomes a little bit too early.
I think that What happens in those moments are we have 15 different sets of requirements all at once, and that's really hard to go from zero to one, from an idea to something tangible when there's that many people. So I think the easiest way to do that and to not lose momentum and still keep people in early is get the smallest amount of people in a room that you can quickly. Someone that really knows the problem well, and someone that can build a solution pretty quickly, get an idea to something tangible.
Sometimes that's the same person, sometimes it's more than two people, sometimes it's three, four, five. But ideally, like build a thing, come out of that room with an idea, and then share that idea to people, and then get all of the different input from everyone so that they have an idea of what you're trying to do.
Because when everyone's stuck at like blank page, then it's too many things coming all at once, and that's what slows things down. So to me, it's like get that small group in, build something that people can react to, that people can respond to, that people can add to, then you move through the regular process.
Kristiana Corona:
I like that a lot because I think it's also not realistic to have too many cooks in the kitchen at the very beginning. You just, you don't even know if the concept's gonna work, right? Like, you wanna be able to make several concepts and see, uh, is there anything here that's valid. But what's interesting is, like, when you go into a situation and, and you're getting stuck, or you feel like you're just continuously hitting a blocker, and you realize it's because certain teams see themselves on the outside, and you can consciously make that decision to change that and to bring people in to curate beta tests.
Like, I remember when I was at Amazon, there was a group that was trying to test a pretty cool technology near a warehouse, but there were a lot of safety concerns. There were, you know, operations concerns. There were all of these things that were happening, and typically, they were not involved in any kind of ideation.
I don't think they had ever been involved in ideation or invited to the table, and when they were, all of a sudden, different options became available. Like, "Oh, we can't do that, but we could do these three other things." It may be eighty percent of what you wanted, but now suddenly we're advocates of this.
We're excited about it. We wanna know what the results are. Like, we're all human, right? And I think finding what strengths that other people bring to the table, especially in those places where you may not know the cliffs that you could fall off, and it's really great to have someone who does.
Andy Vitale:
Yeah, absolutely.
And the, the mindset is so important too of people that really are like, "Yes, and," or like, "I understand what you're trying to do, and here are the risks that I see, but here's how we can de-risk that," rather than like, "Nope, I only have one day to make up this decision before it goes live, and there's too much risk this late in the game, so, like, I'm gonna, you know, dig my heels in on this one."
The earlier you have people in there solving the problems that you don't have to think about.
Kristiana Corona:
I wanna dig into another topic that you brought up, which is how you work, and I know a lot of people are going through the discomfort right now of how quickly their roles are blurring together. So You know, product folks who are thinking about, um, engineering and designers who are vibe coding or in cursor and engineers who are designing using AI and, you know, everyone is kind of empowered in a different way now, and so the roles are really blurring.
Um, and you mentioned that part of that process, that last 12 years-- 12 years, 12 months of delivering some of these things and even less has been about changing the way you work. So can you talk a little bit more about what you're seeing about the way you work? And then also, is there something to abstract out to the rest of the industry?
Andy Vitale:
So I think we're really hitting our stride at the way that we should be working right now. I think it takes time. I, I think in the beginning there's the external, like, everything's changing. What are we gonna do? Everyone's gonna take each other's jobs, and then there's gonna be no jobs, right? But then you kind of like take a step back and you think about it and you say, okay, like, what is the material that we need to build products in, like that actually gets touched by customers?
And that's code. And to me, it makes sense that what we build in is the material that we put in front of customers. Now, designers, some of them know how to code, but most of them don't, right? And I wouldn't expect most of them to write production code. And along came vibe coding tools, and it's able to get people sixty, seventy percent of the way there.
I don't think that just because you have these tools means you can push production code based on different environments that you have and what that looks like, but you can do discovery really well. Same thing with product managers, right? There are gonna be some that are really good at writing code, some that are really good at generating designs, but for the most part, they're gonna get an idea to something tangible that can be tested.
Now, as you get more comfortable with these tools, it's probably okay of like, "Hey, you know what? I have this button and I want to round the corners a different pixel radius or whatever." Like, it's real- it becomes really easy when you understand the environment, when you understand how to clone a repo and branch and, you know, submit a PR.
But really, I think what we're trying to do is democratize parts that would have taken longer that someone else can do. So rather than get on a roadmap to change a button style and wait for that to get prioritized, if you can just go ahead and make a couple of changes and do it, like great. But you have to understand that if you're not capable of doing that, the engineer at the other end of it doesn't wanna fix things that you've made worse along the way.
So I think the safest place to, to do this way of working is through the discovery phase, to get something that feels as real as possible that you can test and do it in an environment that you deploy to, but maybe it's not production. It doesn't risk you messing up other things until you get more advanced to be able to do that. So I think right now we're at this place where our discovery is open to all, right?
Anyone in the company that has an idea should be able to build that, and we have tools to help people do that. We've got tools to help people ideate. We're working right now on how we could democratize our design system so anyone in the company that's ideating can ideate with what our products actually look like.
At some point, how do we get it to a place where we can deploy to the right environments, no matter where you are? That's another thing that we have to explore and see how far we can take it. But the goal is to be able to go from idea to something that we can test and get feedback on as quickly as possible.
Our goal today, too, is how we deploy continuously and be able to ship quickly and get products in customers' hands that are truly our products, not just prototypes of our product.
Kristiana Corona:
Yeah. I think what you're getting at is the gap between what people perceive as being a real finished product and what an actual real finished product is.
For example, all the use cases of different ways it could be used, and d- do we think of every edge case where it could break, and will the errors, and all of those things, right? Like, that's all the non-fun, non-sexy part of this that people don't wanna think about, and they're like, "Well, can't you just, can't you just bridge that gap and just use the prototype and then put it out, and it'll be good?"
And there's still that significant work there that needs to be done. So, is there anything different that's happening to bridge the gap between the discovery work and the prototyping and the final code with what you guys are doing?
Andy Vitale:
I think what's different that's happening is We're not waiting for different teams to hand off items to each other when we're in that process.
I think the days of, oh, the product manager's gonna write the requirements, and then the design team's gonna build a happy path prototype that, like, we'll put in front of people, and they'll tell us, "Yeah, they would use it," but it doesn't throw errors, or it doesn't understand states in the right way. So someone's like, "Yeah, this is great."
Then you ship that to market, you know, months later, and people aren't using it, and you're like, "Why didn't they use it? They said they would." It's like, no, 'cause it didn't work the way that it was intended. So the quicker you can get something that feels real, that you can manipulate, that you can say, "Oh, this didn't work.
I see this error" or, "I see where it got stuck," like, that helps you bring that product to life quicker. So anyone is coming up with ideas, anyone is really working to get something tangible, something working, real, what we'll call working software, maybe not production quality. Then we're bringing the other people in for perspective on it, like, "Hey, PM, I just built this," right?
"Does it solve the problem? Is it the right problem? Are we going after that?" Or, "Hey, engineer, is this feasible to build?" Like, we've done this, but obviously it's not connected to our environments, and it doesn't have auth set up. Like, how would we make that happen? How would we make that work? Or PM on the same side as like, "I just built this, but it doesn't feel like we designed it.
It just feels like I'm getting feedback on does this functionality make sense? Designer, can you look at this and fix this interaction, or make this feel a little bit more seamless, like our experience?" So it does accelerate the collaboration in a way that it doesn't feel like we're leaving comments in files for someone else to try to decipher, but we're actually hands in the material.
Like you said, we're building the clay, and it's like, wait, I've got to help shape this while we're still building it and putting it together. And I think that way just... It gives us an appreciation, a shared understanding of the problem that we're solving and the pain points that our customers are facing, and it helps us all feel like we're being part of that solution.
Kristiana Corona:
You know, it's funny as you're talking about this way of working, right? Because I think about Marty Cagan, and I think about some of the visionaries we've been hearing about for years, and they're like, "This is really how we should be working," and, you know, how the most efficient, agile product, um, ways of being together and making great ideas come to life.
And it-- what you're describing is really, it sounds like the manifestation of more of that way of working. The way we've always known is a really ideal way of working, but now we're actually forced to do it because, like, sitting with the old way and using these new tools, that wouldn't work.
Andy Vitale:
Right. And I think the old way used to produce a lot of Software, but also a lot of debt over time.
And we would have to wait till we launched something to see if it worked, and it would take months and months of effort. And now what we're really doing is we're spending our time thinking about, is this even the right problem to solve, right? Because if we've got hundreds of people all building solutions, there's gotta be a few folks that are like, "Wait a minute.
Like, it's cool that we can do this, but it doesn't make sense to do it. It's not gonna solve the problem for our customer." So we have to think about that and how we think of our end-to-end experience and what we want it to feel like and the seams and how we make it seamless when you go from, like, onboarding with the product to, for us, like filing your taxes.
How does that feel like one coherent, like, TaxAct or Drake Software experience? And that, like, is a very powerful place to be able to say, like, "I'm curating this experience based on a lot of people's ideas," but not everything is something that we need to be building. So how do we identify early the right things and continue to shape them to make them ours?
Kristiana Corona:
What I love a lot about the way you approach this and the way you think about this is it feels a lot more like service design. It feels a lot more like being a facilitator of the right conversations, of bringing the right people in the room, looking at the ecosystem, looking at the best long-term impact or downstream impact for the customer, and you're making strategic choices about impact.
It isn't about, I really love this design or this craft or this, you know, the way-- Now, not that those things are bad. Those things are really good, and you wanna have an excellent experience. You don't wanna have a mediocre or poor experience for a customer because it doesn't look good. But, but y- it's elevating the role of the thinking in the process to the next level and to almost zoom out.
If you haven't before, it's like, well, now you can really zoom out and look at that broader level of impact.
Andy Vitale:
There's a couple of things that Are very new to us, I think as a product building industry now. I think in the past, we used to be so tied to timelines and how long things took and how to get them perfect, where now if it's a bad idea, you didn't waste months of people's time and thousands of dollars.
You wasted a couple of, like, AI tokens and an afternoon. So it's okay to, like, get things wrong. But what's really important too is it's how do we get these ideas to see if there's something there quickly, and then we can apply later on once we know it's the right problem that we're solving. That's when we can apply the things that differentiate us as a brand.
That's when we can focus on things like craft that in the past we would start there because that was the anchor that we had to... I-if it's not a great experience, like, maybe they won't notice, right? Maybe it'll be so beautiful that it'll seem seamless, or it won't feel as painful. Where now it's like, how do we truly make it feel like they can't live without it, or they don't even notice that it's there?
Then we start to layer on the like, well, wait, how do we anchor it to who we are so that they have that notion of, like, our brand and our differentiator? Because software for a little while became very homogenous, right? Like, oh, if you're gonna chat, it's on the bottom right, and if your navigation is, it's gonna be on either the top or the left.
If it's a SaaS product, it's on the left, and does it collapse or not collapse? And all of that started to, like- Swap out the brand logo, and everything felt the same. And now we're trying to get to a place where, like, let's solve the problem first, but then let's also anchor what makes us a differentiator and who we are, and that trust that we have with our customers.
Kristiana Corona:
Sounds like, uh, you're the disruptor who's like, "I am not gonna have the search bar at the top. I'm gonna put the search bar at the bottom."
Andy Vitale:
Well, I definitely wanna stick to people's mental models. But, um, it's always okay-- Like, I, I think that as we think through the future of interfaces, it's crazy. I remember my dad was talking to his phone one weekend that he was visiting me, and I was like, "That's such a weird thing."
And I thought it was because he didn't like type-- like he wasn't great at typing on his phone. But now, like fast-forward a couple of years later, like everybody's... I, I talk using Whisper Flow. I talk commands at my computer all day long rather than type them in because it's just more natural for me. So when I'm looking at an interface, a Claude interface that's just like a prompt, a couple of input fields, feels so much more natural than me like going to a search bar and trying to figure out what I'm looking for and filtering things.
Like every now and then I have to do that, but I feel like interfaces are going so more conversational over time that it's just like we should try to declutter where we can, but we also have to keep information architecture in mind and hierarchy of information. So it's a delicate balance. But I think the quicker we're able to surface what people are looking for more naturally, then it's gonna be easier for them to use our products as well.
Kristiana Corona:
And to your earlier point about design systems, I mean, there's a real possibility that a lot of the work that design and other folks will be doing in the future will be putting some sort of functionality into a completely different AI or system or, or agent, and that interface will be a series of design components or something like that, you know, like Lego, Lego blocks.
Um, or maybe there just won't be an interface because, to your point, it could be spoken. So I know you recently wrote a Substack a little bit just about the topic of good taste and craft and how there are people who are kind of putting their head in the sand around AI because it's like, well, yeah, but I'm a believer in craft, and so it's either/or, right?
Like, I'm not gonna subscribe to that because it doesn't align to the philosophy. But I like your kind of contrarian take on that, and just Craft is very important, and it plays a role, but also there's this whole other world of interfaces and places that you could have value, and you could explore, and you could have an impact that have nothing to do with that.
Andy Vitale:
Yeah. And, and you know, you mentioned design system in the first part of that question, and I really see the design system as the moat for a lot of companies, a lot of the products that we're building, because that is how we take the best practices that we have and our brand elements and our look and feel, and it's how we put it in everyone's hands to help them build, and it's how we make sure that there's a consistency and a quality that's been tested, that's accessible, that's going to make sure that the products that we build s- you know, hold up to the standards that we've put in place for it.
But you know, at first, the people talking about craft were the ones working on really cool consumer products and people were getting craft kind of confused with UI, and then we kind of moved that conversation to like taste and judgment and all of those things are valuable. And all of those things are things that today a lot of the AI tooling doesn't quite get right.
But it's getting smarter, and it's getting better, and it's understanding best practices and we're introducing more skills and figuring out ways to create harnesses for design like we have for engineering. We're not at AGI, right? So at some point it's not yet making the best decisions on some of those things.
But I do think, like I look at my ability to tell when something looks good or when something behaves in a proper way, it's a level of rigor that we've developed over years of expertise, and we still have to apply that, and it's still very valuable. But if we're using that as a way to defend our position, that's just coming from a place of like digging our heels in the wrong area.
I think we need to embrace new technologies when they come into existence. I mean, I've been fortunate enough to have seen multiple technologies get introduced over the years, including the internet, right? Going from internet to mobile to whatever else has come up and introduced. And now this, this AI, it's here.
I think there are some contrarians still that are like, "I don't know if this thing's going to be what we think it is." And the truth is, we, we don't know, right? Like, at the end of the day, we may be moving so fast and accumulating so much debt that we are gonna over-index back on, like, more curated, more, like, personalized...
Well, we're gonna have personalized experiences, but more bespoke experiences. But for now, I think the worst thing that you can do is tell people to try to, like, avoid it or lean into, like, the wrong areas because it's here and companies are adopting it, they're embracing it, they're pushing for it. And if you're one of the people that's seen as, you know, more of a speed bump than an accelerant, that's gonna create a problem for you.
Kristiana Corona:
Yeah, that's a good, uh, good metaphor on that one. I think too about companies that maybe just haven't taken the time in the past to create a design system before or any kind of reusable components. And so people have just gotten away with making it from scratch every time or doing lots of different iterations, and that's really gonna catch up with them now.
And the people who have the design systems are gonna be able to accelerate a lot faster. And I think for the designers in the crowd, uh, the ones that have the design system are gonna feel a lot more confident in the people who are ideating outside of the design, um, or that, you know, they're gonna put together ideas.
You might as well make the ideas look a lot better and be on brand than have it, you know, be clip art kind of things put together or, or AI that doesn't follow the brand. And so it's in everyone's best interest to at least have that foundational set of tools available. I was also kind of curious, besides design systems, obviously, there's many tools, but are there recommendations, things that you're seeing work really well?
I know you mentioned Cursor. Is there anything else that you would recommend? Like, if people were thinking about getting their teams up to the best practice of where we are right now, um, any, any insights?
Andy Vitale:
I think for me, like, I really enjoy using Claude for almost everything. Cursor is-- We actually use anthropic models through Cursor, so we're still technically using Claude at times.
Um Other tools we use, AI-moderated research tools from time to time. Uh, I found them to be somewhat valuable in places where we have very limited research capability. So we use a tool called Genway, but there are a couple of others that also do similar things. We're starting to think about synthetic users because our business is so seasonal.
How do we mitigate some of the risk of some of our decisions if we can start to mimic some of our natural traffic or understanding different customer demographics and behaviors? Can we start to, like, validate designs that way? Because it's really hard to try to get a hold of a tax preparer in May and June, right? I think that- They're on
Kristiana Corona:
the beach. Good luck.
Andy Vitale:
Yeah, exactly. So for us, it's like, how do we start to at least start to prove out some ideas a little bit more than a guess? And then when we're ready to talk to them again, we can see if that's doing the right thing for us. Other than that, I just, I embrace all of the tools.
Like, I love Notion. I think it's really good to store, like, notes for things that I can go back to and sync in one place. I think all of these tools are starting to feel like, through MCP connectors and anything else, that it's able to easily communicate my intent or help me gather my thoughts and ideas and, you know, the Figma MCP can pull in some things that we have that we know of.
So I'm trying to find ways to create my own ecosystem of my tools and my thoughts so that I can do things a little bit faster and a little bit better and more thorough so that it's not like, "Oh, I remember I said this one thing six months ago," right? Like, maybe it's in a-- my-- on my phone in, in a million notes folders.
Like, how do I even find that? But if I can point AI to a Notion and just say, "Hey, like, I know I said something about this. What did I say and when?" And it's like, "You mentioned this eight times, and here it is." Like, that's gold for me.
Kristiana Corona:
Here's the day you said it on, here's who you said it to- Right ... and here's the context around it, so you remember what's important.
Oh, and there was these five other things that you also- Right ... said that you forgot.
Andy Vitale:
It's, it's so good.
Kristiana Corona:
So as I think about the leaders in this audience, um, they're leaders who are across different tech disciplines, right? And, you know, it can be kinda tricky for leaders feeling a little bit in that middle ground of, okay, the people on the team are having to really step up and learn.
How am I keeping up as a leader? Like, how am I going from this place of just being afraid of being replaced? Because, you know, there are lots of layoffs, uh, jobs are being replaced in some cases by AI. But how do you, you know, not fall behind but really embrace becoming fluent in AI so that you can lead your teams well?
Andy Vitale:
I'm thankful to a lot of my group chats that really help me understand, like, where things are evolving. So if I didn't have a chance to see what's happening today in AI, I-- at the end of the day, I will have, like, five or six links of like, "Hey, this is new, this is new, this is new." And then my genuine curiosity of not wanting to be left behind.
I've had leaders that I've worked under in the past that I'm like, they don't feel as in touch as you would want them to be about things. Meanwhile, you know, you find out they're busy, like defending your team in board meetings and things like that, that you're like, wait, of course, they don't have time to know what the latest design trend is, and they rely upon like the team to bubble that up.
So it's more of a bottom-up approach. But I think with AI, we're starting to see a lot of the mandates started top-down. So the team felt like it's being prescribed to them, and how do I stay on top of this? 'Cause every time I look, it's always changing. It's a different day, it's a different opportunity. So I think for leaders, like you've got to extend that curiosity because now is the time where as we see a little bit of compression from the top-down, bottom-up, how do we start to facilitate these conversations about AI?
It's easy to see if someone is just saying words that they read on the internet versus having spent some time in the tool. So I, I think now is the moment where like you've got to just roll up your sleeves and become a little familiar with it and play with it and see what it can do and understand its capabilities because it's still new to everyone.
So the experience that leaders have can help. Like, once you see the capability, you can guide your team to do it and use it in a better way, and you can help them unlock things that they didn't realize were, you know, possible. I think it's really, like, you just have to do it. The more you look at it, the more you read about it, it doesn't help until you're doing it.
Just spend the time, make the mistakes, be able to communicate what it is, and, like, in this new world, like, if you're not doing that, it's a problem.
Kristiana Corona:
Yeah. Well said. I'm very curious, how much time do you commit to doing that, to being-- rolling up your sleeves, being in the code each week?
Andy Vitale:
A lot. Um, uh, you know, I, I feel like for me, I also lead our AI Center of Excellence, so I'm working with fifteen different departments on what their teams need i-in workflows where AI has, uh, opportunities to help them.
Just today, I spent some time looking at our personas and trying to create some sort of skill behind them so that we can start to understand and capture who they are and bring that into the code. Yesterday, I was working with the brand team and took a look at... You know, they have a brand book, and I was like, "Wait, can I turn this into a skill so that we can make sure that anytime we build something, it has the brand rules built into it?"
And that didn't take me a ton of time to do, but it's one of those things where like, how do I help streamline that? And then we have an internal tool where we can link all of our different skills and our different MCP servers. So I'm like, "All right, I created these skills. Let me get this there so that anyone in the company that wants that can access it."
So I spend a lot of time reading different use cases for AI and different PRDs and making sure that things are working the way that they should. So I would say that I'm probably spending seventy-five percent of my time in AI-related functions, and the rest of the time is like You know, working with my peers, helping evangelize what the design team is doing, and making sure we're solving the right problems.
But AI is pretty much in everything that I do throughout the day.
Kristiana Corona:
I would say that is a really good example of the shift in thinking and the shift in time and energy that a lot of leaders and a lot of managers are making right now, where maybe you could get away with not being in the craft and hands-on before, and you could kind of manage at a high level.
But your description, I mean, if you're at a VP head of product design level and you're not hands in a certain percent of the time, I mean, that's a lot of time. But that is the type of leadership that is needed right now. If the team is also going to be spending all of this time, they're gonna be looking to you to-- for direction and how do we do this better.
So getting that familiarity is super important. I love that. I love that you have seventy-five percent of your time in there. That's wild, but it's very cool.
Andy Vitale:
Yeah, I mean, I wouldn't say, like, seventy-five percent of my time isn't like doing what people would consider IC work by any means, right? It's really helping people navigate, but being in the tools enough and creating the things when I see that they need to be created that other people can use.
But I also don't think that it's as uncommon, at least from the people I've been talking to recently that work in very large organizations, to hear the transformations that they're leading their teams through firsthand of like, "Wait, you've moved all of your designers out of Figma completely and into coding tools?
Like, talk to me about that. Help me understand that." And they're like, "I'm in it. I'm in the code. I'm in those tools. I'm, you know, looking at what the team is doing. I'm shaping what's coming out of it. I might not be writing the code, but I'm looking at the outputs on a daily basis." I'm like, "That's great."
Like, I don't know, in an organization that size, that you're doing that is great 'cause you hear the opposite a lot.
Kristiana Corona:
Yes. Yeah. So I mean, that's a big shift for people to think about. Like, it's not that you have to go zero to one hundred and eighty tomorrow, but it's a pretty sizable shift in how you use your time and how you prioritize.
And in a way, it's, it's your new strategy tool as well. So if you're doing strategic thinking and like, what about the future? What about the vision? Where is this headed? How could we be more disruptive? That's the kind of thinking that leaders really wanna be able to do anyway, instead of being caught in the weeds.
So, uh, it kind of forces the function of the right kind of thinking.
Andy Vitale:
Yeah, and I feel like one thing that AI has done is really democratized some of what we know, our knowledge of things, where there used to be a level of like Domain expertise that now is pretty easy to overcome. It's pretty easy to dig in and understand problems in new spaces, and then you apply your lens of, like, how do you take what you know, synthesize that quickly, and then turn it into a strategy that you can use across your organization?
Kristiana Corona:
Well, I know that we're at time, so we got to wrap things up, but I wanna just say thank you so much for taking the time to come in and talk to me, and really talk to all of us about kind of where you're at with AI and what you're seeing from your vantage point. So if people are interested in following you, following along with your journey, seeing what you're building, where should they find you?
Andy Vitale:
Yeah, LinkedIn. LinkedIn's the easiest place. Send me a note. I used to be able to respond same day, but now it's probably, like, weekends. You'll hear back from me on the weekends. But yeah, reach out if there's any questions, and thank you so much for having me on. I really enjoyed our conversation.
Kristiana Corona:
I love it.
Thank you so much. I hope you enjoyed this episode with Andy as much as I did, and if you want to subscribe so that you don't miss any episodes, you can go to worthytolead.co/subscribe. And we'll be sure to include all of the links and resources in the show notes for you, so you can follow up on some of those great suggestions that Andy had.
As always, keep showing up, keep doing the work that matters, and keep leading like you're worthy to lead, because you are. Bye for now.
