Design Unfiltered: AI Tooling Adoption with Allan Ziolkowski

Key Takeaways

• Build a “design sandbox:” Recreate a lightweight version of your app (no auth, no heavy deps) so designers can prototype features in real code, fast.
• ‍Start with your design system: If you don’t give AI your components + rules, it defaults to generic templates and everything ends up looking the same.
• ‍Treat AI like a junior designer: It’ll move fast and guess; your job is to give clear constraints, examples, and feedback so it stays on-brand and consistent.
• ‍Don’t chase every trend: Alan’s move was stepping back from the hype, watching what real teams do, then committing when something actually clicked.
• ‍Design still matters most: Tools blur roles, but taste, user empathy, and a strong point of view are what keep design valuable (and keep work from being “good enough”).

Full Transcript

I'm Erica Borking. I'm joined with Jared Treadley, Joanne Weaver. We are design recruiters based out of parts of New York City and New York, upstate, I guess. Then we are also joined today by guest, Alan. I don't wanna say your last name because I'm going to butcher it. I'm sorry.

I should How do you pronounce your last name, Alan?

Zielkowski. Zielkowski. Yeah.

Alright. There we are. Alan. And he is up in Canada. So thank you for joining us.

I'm I'm excited for today's topic for sure.

I think it's very timely. Yeah. I think it's really interesting.

But like always too, the format of this this little stream is, like, definitely ask questions if you have things you wanna know or contribute or questions for Alan in particular. That's why we we bring people like Alan onto this and and share so that we can create this open forum. So really, I'm excited for this topic around AI and tooling and designers. But, again, you got questions, let them rip. That's what we're here for.

I'm excited because, you know, we did a little bit of a different thing a few weeks ago. We posted a, like, a what is it called? A open call kind of for anyone interested in being a guest. And, Alan, I believe, was the first one, and you were so excited in your entries that I was like, we gotta have this guy. So, yeah, I will give a quick little overview. Alan is currently the head of product design at Beyond Identity, which is a cybersecurity startup that is focused on passwordless authentication and identity security.

He spent six years at Apple working on AI machine learning tools and was also the first design hire at GlossGenius.

He's built his career at the intersection of product thinking and emerging tech. Deeply passionate about the future of design, which we will experience today. He has recently helped his team become truly AI first, and we're really lucky to have him today to share kind of what that transformation looks like in practice.

So, yeah, any questions that people have, pop them in the chat while he's talking. We'll pull them over as usual.

But thank you again, Alan, for being here. We're really excited to learn more about this. Yay.

Thanks, Alan.

Thank you for having me. You, guys. And thanks for doing the open call as well. I'm giving a platform to be able to talk about this.

Honestly, it was a really good move on my part. I'm just gonna say that.

Love it. Full credit.

Go, Erica. Yep.

Full credit.

And by the way and as we're talking, so for our guests, you know, this is an open forum.

This is a livestream. We're all live. So, you know, as you come in, say hello, say where you're from. And then as questions pop up, as soon as you have them in your mind, put them in the chat, and you guys can talk to each other too. Like, this is all a forum. It's a community.

We love people cross talking and, sharing ideas, so that's what this is for. So if you have questions for Alan, just put them in as you think of them, and we'll get to them when we can.

Awesome.

Where where in Canada are you, Alan?

I'm in Ottawa.

You're in Ottawa. Yep. There you go. Well, that's some good variants we got. I'm seeing Denver, Minneapolis, North Carolina, San Francisco.

Yeah.

Good to have some Canadian representation in there as well.

Appreciate it. Yes.

My do Alan, do you wanna kick off with, like, a little, I don't know, icebreaker questions since we're gonna be talking about AI Yep. Tooling.

Yeah. So I think lots of pressure, lots of FOMO happening out there right now. For those working at start ups and and on design teams, even freelancers, what are you feeling right now?

Where are you leaning towards? What what, what types of pressures are you getting from your team to kind of incorporate this, and where are you leaning, towards getting your info? I'm really curious, because it's something that I went through, with my team over the last six months. You know, the AI journey started more than a year ago. For us, we're starting to get a little bit more serious, but I am really curious. Like, you know, now feels like a really interesting really interesting time in design related to the agentic world where I think is is changing a lot of the game to enable us to do faster, smarter work.

Yeah.

So it's a huge thing. You you know? I feel like we're hearing a lot about this, changing the game, interesting time, wild time to be a designer. Like, this is the the topic of the of the of twenty twenty six so far, I think.

And I think there's just so many people too that, like, don't even know where to get started. Like, there's so much signal to noise ratio out there. Like Yep. Like, I love when we had our little huddle yesterday getting to know you, Alan, and you said one of your designers didn't know what GitHub was two months ago, and now she's in the thick of it.

Like, she's doing all this cool stuff with, you know, designing straight into code. So it's like anybody can do this, and I think that's what I'm really excited about for this conversation is that I think it could be really orienting and really just empowering for people to know, like, okay. What is this? Like, what do I need to know?

And then how do I get started? So Yeah. Definitely. That. Yeah. Yeah.

Before questions jump in, I I'd actually I'd love to talk a little bit about, like Yeah. The history of how we got here. Yeah. As a place, I think it's really interesting.

I think, you know, the this this thermal pressure or, like, how do we actually incorporate AI has has been really hitting for the last, like, a year or so in picking up picking up. And for my team specifically as a design leader, I've always wanted to try to make sure we're at the cutting edge of what we're doing and the and the right thing. But the the examples I was seeing online just weren't matching to my environment to what I actually had to try to do. Right?

And I'll go into that a little bit more, but I think, you know, if we look historically about a year ago, I remember, like, from our angle, design's input into AI was really about, like, let's get this AI feature into our tool. So how can we think about using AI within whatever product we're building? There's some chat, some you know, in ours, like a natural language policy editor, things like that.

Right? Yeah.

It was, like, kind of interesting, but everybody was just kind of talking about it. Some people were jumping on it. Some people were doing things. It wasn't necessarily the most, like, intuitive thing.

And we jumped into this phase of, like, leveraging generative AI and things like ChatGPT for doing some research and cleaning up Poppy and things like that. And that kind of happened early on. And, you know, I'll give my timelines. People everybody's in a different stage of where their adoption Yeah.

Be really clear. So you might be at that stage right now and kind of trying to figure it out, and, I'll share kind of the the like, this is not something that happened overnight. Over a year, all these things transitioned, and the last two months is really where we made that that commitment and that leap. So you're talking about, like, using ChatGPT to clean up copy and things like that.

And we were doing, like, generative imagery. So, hey. Now instead of, like or in product design, can we generate, images to go into our onboarding and things like that? Like, that was, like, maybe an interesting way to do it.

I was always kinda looking at it. Was like, this can't be it. Right? Like, this can't but it was still, like, helpful, and it was still cool, and I was still leaning into it.

And then, you know, two things kind of happened. So the generative phase happened. We were bringing things from the generative world into Figma, and we're kind of tweaking it and doing that, like, crossbreed thing. Two things kind of happened.

The first thing is, like, something like FigmaMake came out where it got really interesting to build these prototypes, really quickly. And I remember actually doing it for, like, some demos, really quickly and spitting stuff out and being pretty impressed with it. That happened, and that was like again, it was like, okay. This is kinda cool.

But there was one thing that happened in relation to that that always kinda made me feel a little bit uneasy is that no one ever actually asked me for the code that was being generated from FigmaMake, which was really really interesting to me. And that was, like, one of the first kind of clicks that appeared.

The second thing is that the agentic world started to pop up and started to kind of explode.

So FigmaMake in relation as an agentic tool to building UIs, but other people started to kind of grasp access to this super fast. And that was cool and terrifying as a design as people were just making their own apps, really quickly, and you were, like, trying to follow along and see what's going on. And really realistically and this might be happening at your company right now, like, leader, whoever it is, building stuff, with no design guidance. And one thing that I had noticed actually early on was the the project itself needs to start with some baseline foundation.

Really, that is design system. And these people who are building this software without the knowledge of the design design system, we're just doing stuff, the AI would say, okay. Like, hey. I'm just gonna recommend using this, Chad CN or Tailwind or whatever Design system out of the out of the box space.

And that's why everybody's stuff looked the exact same. Right? You got these purple gradients, we got this world that it was, like, called AI slop UIs because they didn't know. Right?

And designers weren't really leaning in and jumping into that portion. So that was actually my first venue into a, like oh, okay. First of all, super scary. Things are actually happening in my world now related to this.

It's actually affecting design and my space, and I need to try to do something to to control this here. And I don't really know what it is, but what I knew was, okay. We have a design system we're using already within the company, and I'm responsible for the brand design as well. So this was kind of leaning from the product as well as the brand.

And I wanted to lean in and just be like, we've we've gotta do something here where we've gotta maintain some some bolt on control. And so I just, one night, just whipped up Cursor and took the foundational libraries that I knew that we were using and just started prompting it to build out a design a new design system that I had full control over. Because I think that was one of the things that was happening at the time is that the design system that our company was using was tied to the enterprise code base. And it's complex.

It had a lot of dependencies, a lot of things that were like, made it a little bit more difficult. So for me, it was just like, can I build something out of the box here? And that's what I did. I built a design system out of the box.

And this is early enough that, like, we weren't leveraging things like skills and CloudMD and whatever, where I had it make a website, essentially, like a storybook, an output, and I had it spit out a bunch of AI rules on it, and I had it do a download, for a package. And so what would happen here is, I would say, like, package this entire design system that I've made over here and embed these rules as part of it, and I would give it to someone like the CEO or whatever, whoever would make the tools. And I'd say, when you start your project, download this package and put these instructions in there, and hopefully, it'll reskin.

Right.

And it kind of did, and it kind of did, to be honest. Right? And and we could pause there because this is kind of the a little bit of the journey, and we can talk about, like, where we actually get into the real stuff. But that's, like, a little bit of the gap or different steps that we we took through.

Yeah.

I mean, you're you're diving in. I mean, you saw you saw the need for it. Right? You saw this this gap where, like, things are changing, how the way of working has differed now.

You know, our companies change you know, these stakeholders, these other counterparts are changing how they're doing their work. Yep. They're kind of kind of coming over to design territory. Yep.

I think a blurring of the lines is okay for the most part, but, you know, designers still need to to design.

Yep. It it the the it sounds like from your story too, there was no, like, resistance around, no. No. No. Stop everyone. We gotta keep it the old way.

That's that's it. Right?

Like, on.

Yeah. A big part of it is the ability to lean in and and Right. And that, like, if you don't, you'll get left behind, especially if you're working at a company that is, like, trying to push that leverage or that lever. If you're the person who is trying to do that in your company, you know, you're gonna have different challenges that

Will come up. But I think, like, in general, especially, I don't know if there's any design leaders, on the call. You might be trying to think. You know, I got a lot of messages when I had announced that we've done this for our organization, from some design leaders of, like, how'd you do it?

Like, you know Yeah. We're putting this pressure that's coming in here. I wanna kinda figure out is there something that's realistic. And that's one thing that's really important for me is I didn't wanna just jump into the trend or figure out what the next, like, cool tool was to use because, truthfully, I was, like, exhausted.

I think we were talking yesterday, and I was I was saying that, like, I literally logged out of LinkedIn for a little bit. And Yeah. I stopped watching these YouTube videos of, like, this is the next best thing or product designers are dead. The role is gone.

And I was having these, like, conversations with my my team, one on ones with the designers. Like, what is happening, and what are we doing here, what are we gonna do? And a conversation with a friend outside of of work.

He kinda told me, like, just pause. We don't know yet what's gonna happen. I think that's okay. It's actually a good thing to be able to say you're not just jumping head first and taking a breath and kind of looking at what's out there.

And when something connects and clicks with you, that's when you can start to make your move and progress. And so that's what I did. I took a little bit of a step back. I stopped Kind of feeding into all the hype and tried to just, like, understand what's going on out there.

And the other thing that I talked about that I think was really important is not just, like, looking at what influencers or content creators are posting about this. Looking at, like, what people that were working in the industry are actually doing. Right. They're actually talking about and that's what I wanted to do with myself is, like, actually share a little bit of the story of, like, in an enterprise environment, you can make this work, and this is how I'm doing it.

I'm not selling you anything. I don't have any any motive other than trying to just share my perspective through my experience of how it works for my type of environment for the Yeah. I was in. Right?

Yep. Yeah.

I think that's such a healthy approach that you did take some time off, that you just kind of cut off the world because it was just giving you a lot of kind of verbal slop, right, to use the AI parlance.

And just to figure out, like, what's really working for you and, like, what do you want out of it. Right? And, like, what what are you connecting with so you can come back and say, hey, everybody. I've had some time in my cave and kind of work with this, and I've gotten my my arms around it.

Like, here's what I think the the positive benefits can be or, like, here's how you can use it. Because I think there's just I think there's a lot of fear out there right now, and there is a lot of noise out there right now. And I think people are feeling like, oh my god. I'm getting left behind.

I don't have it figured out yet.

But I think, you know, from talking with you, it seems like this is something that one can pick up fairly quickly. Would you say that that's accurate?

I would say that, depending on the, type of environment and work they're doing, definitely.

I will say that it's not the answer for everybody, and it won't work everywhere perfectly, and I wanna be transparent about that. It's like, it's not the be all end all solution, and not claiming that. But it is something that you know, doing this for fifteen years was the first time that it clicked. I was like, this can actually make a difference for my team, and it makes sense, and it clicks.

So tell us about that click, Alan. So what was that thing that clicked for you, and, like, how has that revolutionized your design process now?

Totally. Okay. So I wanna give credit, Real Lu from Cursor. Yeah. I was starting to watch a lot of his videos to see what he was doing with his team.

One of the the PM leaders on my team was like, Alan, you should check out what, Cursor is doing with their internal design team, to see how they're operating. And it was a a YC video that he was doing, an interview. And at one point, he started to talk about something, and this is really the meat of of everything. He started to talk about how first of all, he was reminiscing about a lot of the challenges that exist with the existing way of working, that we would build out, spend weeks kind of doing everything that the books told us to do for design, do the research, build up the mock ups, get everything ready to go, build out your prototype, etcetera, except when it was actually coming time to shipping and production, only, you know, twenty, thirty percent of what you had designed actually made it in because of the reality of business, what happens, and how things ship quickly.

And so that was the first thing. Was like, oh, this guy gets it because we're in a similar kind of space here. But the next thing was his introduction to something he called baby cursor.

And baby cursor was the moment for me where the complication of a designer trying to leverage UI and work in these really complex enterprise systems, is really hard. There are so many, dependencies that one needs to try to get across. Right? Even for me to get the design system so I can mess around with an AI was a problem.

There was just so many hurdles that had to cross. And so what he did is he just spent a little bit of time remaking the enterprise tool, Cursor In Cursor so that he had a sandbox environment for his team to play with. And so that's exactly what I did. We call it, we call it our design sandbox.

And what I did is I took our enterprise app, took me less than a week to go page by page and recreate it, with a with a whatever tool you're using. Right? Our tool of choice is cloud code. I was using cursor at the time, as well.

But recreate your enterprise app for yourself without all those dependencies, without the authentication requirements, without the pipelines, without the merge requests, all the things that you need that get really complex when you wanna start messing around with things, and make an environment for yourself. And so what happened is I was able to build rebuild our app. And, you know, it doesn't need to be one to one perfect. It doesn't need to be every single page, but the main functionality rebuild it and apply that with the design system, as well.

And all of a sudden, I had was a repository, main branch repository that my design team could go take that, and when they were working on a new feature, start from that and add things onto it. And so as an example, one of my designers was working on a new devices page for a feature. And what she did is literally took the the baby version of the enterprise software that I did, and I'll talk a little bit more about what's the the mean bones of that. Replicated it, made a new branch of it for herself, added the new page to it, and because of the guidelines, skills, constraints, that we've defined in the agent, she was able to just say, hey.

Let's add a table here. Here's the columns that need to live in the table. Here's what the filter need to look like. Here's what this needs to look like, and build it all up super quickly.

Now she has a working branch or repository that she can share with the developers, share with the PMs, whoever it is, and they can immediately start playing with it without requiring all these extra steps and gave an output that is actually, in my opinion, higher than what we would have done before with Figma. Because let's say your, you know, minimum output is a static block that you're leveraging your design system or whatever. Now out of the box, we've built in accessibility into our design system. We've built in responsiveness into our design system.

And all those things are being taken account for by the agent when that person's designing things. And so you get all those things for free as they're working. States every single thing. And that developer, that PM, that salesperson who wants to demo that new feature can download your repository and just run that demo locally over here and start to see and play with that app that looks and feels exactly like yours except, you know, it's your design space.

And one thing I wanna just add with this, like, in your in your sandbox environment, another thing that often happens is you don't have real data when you're kind of trying to work with things. If you're trying to work with things in in something like traditional ways of of doing design, you have to add all these things manually. In the old way, you know, you're even talking about, like, Lorem Ipsum and stuff. This is using real stuff that you're generating and putting in here.

Not just that, but, like, I can tell it. Let's add a thousand fake customers that, you know, look like this profile. Here's the types of devices they're using. Let's add this to our fake database over here.

Let's add a different variances over here. I want some logic to be twenty percent, failing authentication, eighty percent passing. And then all of a sudden when I built that, like, backbone of that database, this mock data, now I can go and say, let's generate a new dashboard, a new a new chart over here, and it'll pull that same reference data that we're using within this.

Yeah.

Everything is consistent through it. So you have a real feeling experience that I can go and test with actual end users as well. And just to be, like, really clear what what I'm talking about from the enterprise unlock. It's rebuilding your enterprise app page by page, in your IDE and, AI tool of choice and using that with your, with your Git service and allowing the other designers on the team to branch off of that and add things and then merge that into main. You have your own play space now that doesn't affect anybody. Yeah. And that's that that was the ultimate unlock for us.

Could you imagine, like, when you talk about shared language and communication now with stakeholders and even with end users, testing with those end users, the efficacy of what you're testing with now, which is essentially a a real version of the product Yep.

Versus how you might have tested in the past, which might have been a static mock or clickable maybe even a clickable prototype, a Figma maybe. Or yeah.

So clickable prototype, but I think, you know, here's the other thing that was, like, a a problem with that is, again, remember that, like, that percentage of what actually gets built perspective. Right?

Yeah.

What I was showing, you know, it it wasn't actually super realistic to what might actually end up getting Yeah.

Exactly. Right?

So if I was building this out and it was outputting code, you know, this is quite realistic in what a translation could actually look like Right.

And get built. Right? So within those constraints, I think it's it it kinda changes the game from that perspective. And I think that was really, really, really interesting.

This is understand oh, sorry. Go ahead, Jared.

No. Just just reaffirming, like, this is what I continue to hear from other teams, other companies, other designers I speak with. Like, this increased efficacy, this better way of sharing work now, this better way of communicating with your stakeholders, with testing, with getting that into production, It just keep it's just this is better. It's better at it. It's not my opinion versus that. It's like we feel we're getting better results because of doing it this way.

Totally.

And just to add, I know you're you're about to say something, Joanne, but just to add to that, you know, we're working in the world of product design

Where essentially at a high level, you are working with LEGO bricks that the design team has defined. You've said I want these LEGO bricks to be red and this shade of orange and Yep. Whatever it is, Adams, variants that you wanna have with it. And, you know, when you're working with your design system in a traditional way, you're just grabbing those and piling them on top of each other.

Well, instead of me manually doing that now, I'm just telling the AI to take those to define ones that I have, and I'm just putting it, from that perspective. You know, with something like brand design or one offs or things like that, we haven't had that same success, truthful from our side to do that full kind of shift over. But this lends itself super well to AI because it's working off something it knows. It's working off something it can reference.

I tell the I literally tell the the the AI with rules, if you're gonna build a new page, go and reference another page within my system and make sure you're doing things very similarly so that the patterns stay consistent because that's how you build and design software. You don't you don't go and make a a a new page look completely different, operate completely different if it's got the same patterns. Right? So this is what you're doing.

Love it. Sorry, Joanne, to interrupt you.

Well, I was gonna ask a different question, but now I have a different question based on what you just said, Alan, which is, so and I think you said yesterday in our huddle too that that it feels like when you're telling the AI to do this, do that, like, you know, use these predetermined rules or this design system to accomplish this. And in some ways, it feels like you are kind of, directing some junior designers.

Yep.

Right? To make the finished product.

What can you say to people out there that are junior designers now or, like, looking to get into this this industry that might be afraid? I mean, there's, like, that whole trope out there that's like, oh, AI is gonna get rid of junior designers, and their jobs are gonna be gone. What would you say to them right now? Like because they're probably feeling a lot of anxiety and, like, are their jobs gonna be taken by AI? But what's your take on that?

Yeah. Great question.

So here here's what I'll say. Probably a lot of the traditional things that you're learning, in school in at the moment, might not have fully caught up, and there's this weird gap that's happening where we're in this transition phase where no one really knows exactly, where this is gonna land. This method that I'm talking about is an interesting method that we're doing and working with, but there's been a lot of failures and a lot of things that have come out of it, as we're working as well to be to be transparent. So it's not like there is a clear definition or answer for someone that says, learn this, and you will be set up for success.

Right. I think the big thing that, like, I would tell a junior designer is that lean into it, and experiment with it, and be honest with it as well. It's like, I don't think this is an elimination of the role. I think there's always gonna be highly qualified and skilled people that are gonna be able to orchestrate the AI to do things.

And, you know, one example that I give to that that I shared with my team again is that, you can give the, the same kind of requirements to a a PM who might be able to build the UI, a designer who might be able to do it, and an engineering manager who might be able to do it, and a whatever. A CEO. Right? And what they're gonna do and spit that in, and modify and change it is gonna look different depending on their angle or perspective.

Right? And I think the real thing is that everybody is gonna kind of shift and, you know, we're talking about this blending of the world, is that a cohesive product building world is gonna kind of, appear out of all this, but your skill set is gonna be the thing that's gonna lean in and be able to jump in and help adjust the things that matter from a design perspective, the user empathy, what it looks like. You know what good looks like. Right?

Have that, that visual, eye, that capability to see things from a different perspective. And that input that you're gonna give is really how you're gonna potentially blend into this new world. And I say potentially because truth is nobody knows. If anybody if they do, they're lying.

Agreed.

Agreed. Yeah. I I was picturing, like, a Venn diagram where it's like you got the three legged stool. You got engineering. You got product. You got design.

Yeah.

And now the roles are starting like, that Venn diagram is starting to overlap more and more with each other.

Right? But it's like but those edges on the side of, like, where they differentiate all those three disciplines, I think design really has an opportunity here to really be a differentiator because they are, as you said, you know, so user focused, empathy focused. They know what good looks like. They know what good should feel like and act like. Right?

And and, you know, just before we talk about this too, transparently, there will be some companies who will bypass this type of role, and they will Yeah. Go with what's good enough. That's just gonna happen by nature of it. Right?

Yeah.

But that's not a place you wanna work anyways, truthfully. So Yeah.

Totally.

Wanna be in a place that values and understands what design can bring and how it can help differentiate a product.

Yeah. Yeah. Absolutely.

And when you have all of these different people in the company, maybe sales, maybe it's the founders, the designers, the PMs, engineers, like, who ultimately at the end of the day decides the direction that you go that you go in?

Yeah. That's an interesting one as well because the the lines are crossing and the worlds are kind of crossing. Yeah. Again, I think it's one of those things that the industry is trying to figure out and shake out that's happening right now.

But I do think there is this chain of command of, well, this person has this capability to make these adjustments over here and bring in this skill set here. And so we wanna make sure that they're actually jumping and leaning in as part of this. And you look at how people kind of lay out their processes in general, you know, like, hey. We can't this ticket you know, let's look more traditional.

This ticket isn't closed unless designs actually reviewed and approved it. It's kind of the same type of model, right, that needs to kind of act.

We still need to be able to work cohesively together

With your other peers and partners for it. Because what is stopping theoretically someone from going, punching something in, and just bypassing design completely?

Right? But nothing nothing's stopping it unless the teams are working cohesively from that perspective.

Do you do you have a maybe a a quick tip or two, like, as these lines blurred and everyone is a designer, product product managers are designers, engineers can can design. Yeah. As you said, Erica, your CEO can design. They're all coming up with their ideas, and maybe they're of vibe coding things or doing things with their tool of choice. And they're going, hey, design.

Like, we've we've done it. We've solved it. Or here you go. I've come up with the with the new thing. You know? Maybe you guys can just execute through or you could just finish this off.

Like, how do you as designers then, even just from maybe a stakeholder management perspective, like, alright.

Hold on. Like Right. Let's you know?

Yeah. I think it's it's it's the responsibility of of the design leader, at the company and space to make sure that that design voice is heard and brought to the table and understood the importance of that level. Yeah.

It's it's responsibility of the company to understand that that position brings a a unique perspective.

I think, you know, the the scary thing that's gonna be the opposite is if you don't lean into it, they will just move on without you. And that's really the the the thing I want to kind of make clear is that, like, if you just kinda sit back and and and and not do this and they are starting to work these knowledge workers are starting to work way faster And you are stuck in your old ways of doing things, your contribution now is the bottleneck. And that's one thing I always try to put with my team is, like, let's never be the bottleneck in this entire process. I always want design to be ahead and to look at things.

And the traditional model of design is design is at the front of the process, and we come in at the end and kinda check things. When you work in this way now, in this new model, you are no longer just at the beginning. You are literally at every single step of your way. And that's a like, as a designer, the ultimate fulfillment that you're being included, that you're being part of the the solution all the way through, and you're seeing it through.

And what you designed and built and intended to actually get out there is what's being shipped as well, which is amazing.

That's that's gonna feel good going from the I hope that twenty percent ends up in the the final product to kinda being more like, there's no reason why this shouldn't all be there.

Exactly. Exactly.

And the the the thing related to that let's talk about, like, how people bring it into their org Yeah.

In relation to that. I do think the the best way you're gonna have success is if you have engineering partners, EMs, individuals that see what you're trying to do and become allies to you because you're gonna have a lot of questions, and it's hard to figure out on your own. And they're gonna be able to see the benefit and help ramp you up. And I would say there are, you know, a handful of people at my org that really push and allow this to to become true and real, and we wouldn't have been able to do it without them. And I think because of that now, everybody in the company is reaping the benefit of it. But these people that are not afraid to say, hey.

Oh, just try it. Just do it.

Or Yeah.

Here's what you need to do. Or you jump on there and you're like, hey. I'm getting this weird error that's popping up. What do I need to do?

And they're like, let's jump on a call. Let's look at this together. Let me explain it to you. You know, we talked about, a designer on my team who who, you know, didn't didn't know what GitHub was before.

This is something she had. She had a partner in an engineering, leader who was able to see the vision and say, let's make this work. I see it, and I'm going to help you kind of get through this. And quickly, the concepts stuck together and moved really quick.

It sounds like everyone's working oh, go ahead.

I was just gonna say it sounds like everybody is working better together. And Amara said people are appreciating designers and tools a little bit more. Right? So it sounds like maybe it's a little bit more seamless versus, like, oh, engineering doesn't care as much about the design as the designers do.

Right? And vice product and such. Right? Because I think we've always seen kind of I don't know if it's like them versus us kind of thing, but it sounds like everyone's really, like, all excited about design now, thankfully, finally.

We've only been waiting forever. Right?

Yeah. I think look.

Absolutely agree. I think there is a, an element of being a little bit sensitive to what this is doing to some roles that is kind of unknown. The traditional front end developer role is kind of shifting and switching a little bit now, and I think no one knows exactly where that's gonna play out. And, you know, I don't my angle of perspective was I'm not coming in to use these tools to take over what you're doing.

That's not my intention. It's making my team a lot more effective, and it's unifying everybody in a much better way. We still highly interact with our front end team, and we have a couple different processes which I can share from how we move with the sandbox. Right?

Because one of the things I mentioned is we have instances where the designer is shipping stuff into production Yep.

Moving forward. But the other thing is we have the opportunity where we just leverage the sandbox, and we work with a front end engineer who is the expert in that domain, right, and can look at it and say, I see what the AI did over here. There's a better way to do it. Here's how I'm gonna convert that code with cloud or whatever and kinda build it the right way, but I understand the vision of what you're trying to do.

Yeah. And that works really well as well. So we've been doing that model. We've been doing the model where the designer can contribute directly to the code base, and we've been doing a different model too where an EM or PM actually starts.

They build the first kind of iteration of the version, and we jump into it with them and go through the translation of what that experience should be. All three of those models have worked, and we haven't figured it out fully.

But I guess each each use case would have different you know, in some instances, yeah, do model one.

Another instance, model three. Yeah.

You can There's no And what I love about this too is that it just it feels so empowering for design.

Right? Like, I think whenever I'm sure Erica and Jared, you get this a lot too as recruiters, as fellow recruiters, is that, like, when we talk to candidates who are looking for jobs, one of their number one most important things that they want in a new role is I wanna be at a place where design has a seat at the table. It's not just an afterthought. It doesn't report into engineering. Even reporting into product is sometimes a little bit iffy. I mean, the most of the time, it does.

But in terms of, like, just practical, actionable advice, Alan, and I think you're giving great advice, like, you know, make friends with your, your engineering managers.

Make sure that you have a, like, an ally in them. What are some other things that that designers can do, like, to seize this moment and to really ensure that design is going to have a strong seat at the table, that design is going to be a big part of this aside from just, like, getting in the tools and learning them. That's, like, you know, table stakes, it sounds like. Yep. You know, getting close relationships with engineering. What are some other just kind of initial steps people can take just to to make sure that they're, you know, that their job is safe and also that they're, you know, that design is well represented at the company.

Yeah. I I think, Steph, out of the functional world of the doing things Specifically, and think about all the other things that make us designers. Right? And it's something, like, I try to put and embed with my team.

It's like, do you dedicate time to kind of step out and look at what good design looks like? Do you have a good opinion of what good design looks like? Strengthen those skill sets outside of this. Right?

Like, this all we're talking about right now is just tools and processes for how how it's kind of evolving. But, like, if you are not that opinionated designer, it's gonna be a struggle regardless. Right? So you still need to maintain all those things.

And so I push my designers on my team here to say, like, go understand what the latest and greatest is. You now have a new leverage or medium afterwards that you can try to experiment things even more because you're not kind of handcuffed to the tool that you're using. So I think, you know, what I would say is, like, don't let your world just be about design like, AI design and what tool should I use.

Maintain that, like, thing that made you excited to become a designer.

Yeah. Keep that there because now you will become the differentiated designer that pops up. I see a lot of questions kind of popping in.

And Yeah.

Yeah.

I I wanna see if there's anything we should start addressing because I wanna I really like, I really wanna try to help people and answer answer answer genuine things of, like, what what's worked well and what hasn't worked well.

Yeah. I think if I grab one, I feel like Leslie's a little bit of a long one, but there were a couple of interesting things in there about maybe the speed at which people are are moving at, you know, scales of of, you know, companies and their budgets and whatnot around this stuff.

Yep. Yeah. Look.

One of the things I was prefacing is, like, not every company is gonna be like this, and don't be scared of that. Right? Like, there will be many tools and many larger companies that may still operate the same way they do. It doesn't mean you as a designer can't explore and understand how these things work. I actually think we're very much on the pioneering forefront of, like, how these workflows work.

I say that because I can't look up what the right answer is because no one knows what the right answer is. Right?

And so, you know, I guess, Leslie, I wanna make sure I answer your question correctly, but five years later, you're a company. Yeah. This this is something so one of my designers, talked to me about this and said, like, hey. Everything we're doing is really cool and forward facing.

I don't know if my next job will be like this. And I said, you're right. I think if you go and you apply, for a job in the traditional kind of manner, I think, you know, it's very possible. They haven't adapted or adjusted their workflows or either workflow work for them.

I think the people that are doing this type of work, theoretically, are going to be the ones that, are working with recruiters like yourselves that are looking at these cutting edge companies that are trying to fulfill these roles of people that don't necessarily know how to do this, actually in practice or haven't done it yet just because it's so new. Right? So I think, you we are in this weird you know, we are in this transition phase. There are peep people who very much are working in the traditional method.

People are working in the new method.

People working in the new method like myself Don't have all the answers. A traditional method has time time has evolutionized it. You know? When I was working at Apple and I was been doing this for this long that, like, I Photoshop was my tool.

Wow. What I did. Right? And even the the type of work that I was doing at Apple, we couldn't use cloud based tooling, and so we weren't able to use something like Figma even though that's what everybody was doing.

The fundamentals of what I was doing and what I understood translated really quickly and easily, but I was behind on that game. And I think regardless, it's the skill set and the knowledge that you need to know. There was a time when I was at Apple that I was doing designs in high fidelity in Keynote because it was the fastest way for me to get it to a stakeholder so that they can leave comments and even make small changes and tweaks and stuff because the Photoshop way didn't work. Right?

Like, they didn't have Photoshop. They didn't know how to use that complex tool, etcetera. So Right. These tools transition.

It happens. It's it's really about the knowledge how to manipulate and do things. That's gonna be the skill set.

It's a great point. That's that to me is what makes a great designer a great designer. They're they're they're great over time because the tools will come and go and the the ways of working will change and evolve, but, you know, having the the fundamentals or having that ability to adapt and grow, that I think makes a great designer consistently.

Yeah. If I can jump to as well the creative director analogy that I've been using and that I say with my team.

I think the thing that's really interesting about it, and it's like, you know, how do you work with these actual tools once you're jumping into the prompting world is it's it's talk talk to your AI in that sense that the the AI is a really eager A b junior designer that wants to try to get things right for you.

But if I if I were to tell a junior designer today, hey. Let's build this and just give that context.

And because it's eager, it's just gonna try to do it. Yeah. And it's gonna spit something out that is okay, but it's gonna miss a lot of context. Right?

And so if I don't give it examples, if I don't share what I'm trying to do, if I don't give it guidelines and restrictions on things, it's just gonna do what it wants to do. And so I found that actually with our design system, we were leveraging Bootstrap icons. And at one point, I was doing what I said, right, where we're pulling in in our sandbox environment and a different designer was pulling something in, and it got pulled in and it wasn't using Bootstrap icons. And so what I did is I talked to it as if it was like a human, I said, hey.

What's going on? Why aren't you using these Bootstrap icons? And it did the classic, oh, you're right. But it explained to me that, look.

I was trying to go through the path of resistance. In the future, if you want me to do this, like, let's let's set some constraints. And that's that's literally how you have to talk to it. It's like, if it was an eager designer that you're working with and they're trying to get something done for for you, they might go and, oh, there's no icon for this, so I'm just gonna pull it from a different library.

Well, how did how did that designer know that I wasn't okay with that? And that I would actually have rather they messaged me and said, Alan, would you like to use a different icon? I don't actually have this one available. That's really what I wanted to do.

Yeah.

That's what I told it to do. I said, in the future, you know, let's adjust these rules here. Let's adjust the the skills, and let's let's have it say that don't just go rogue and try to do things here.

Come back to me. Let me know why it didn't work. Let me know what I need to do to make those adjustments. And so I think always kind of releverage yourself in that type of conversation to it. And, you know, the other thing is there's two kind of perspectives I've seen work.

I don't know what the better one is, but the two kind of angles is either a, you ask it to do stuff very small. So, again, working with that junior designer where you're like, okay. Let's start with a a sidebar here, and let's add this one page over here. And now we're gonna work on this page.

Okay. Let's add a top bar here. Let's add this to take the header of this. You can work, like, very, very incremental.

The other thing is you can work a little bit more chaotic where you can tell it to do something, and now now you work in the opposite direction where you're like, cool. You did this. You gotta strip this back now. You gotta strip this back now.

You remove this.

You gotta remove this. Gotta remove this.

Is one way better than the other?

It I think it depends on the scenario of how you're working. I think the chaotic way can be overwhelming for designers is what I've kind of seen because the AI goes a little bit rogue and does things that it booms you want. And all of a sudden, if you're now working where as a designer, we didn't traditionally work with datasets and things like that, and all of sudden it's changing logic in your system and changing data and things like that, you you get a little bit overwhelmed. So Yeah.

I like a little bit more of the incremental approach to to kind of nail out each piece, but it is a little bit slower. Right? And you think about it, like, from a human perspective, it's the same thing. Right?

Like, if I'm telling you do this. Okay. Great. You gotta go do that. Now do this next thing.

It's like, it's slower, but it's probably a little bit more accurate.

Have a question. Oh, sorry. Go ahead.

Oh, yeah. Are you gonna bring up Sean's question?

I was. Yes. Yay. Go for it. Yeah. I can actually gonna be more.

I can't put it on the screen some there we go.

Great. So Sean Rosenberger. Hello. You said, hi, Alan. You mentioned this process.

You described didn't work very well with brand design. Why not?

Yep. Great question, Sean. So I think there are tools and things out there that will work well. I will say this, like, design sandbox concept that I think really connected and clicked as something that I'm excited to talk about works well because it works on the basis of foundation of, like, Lego pieces that are being built upon. And so brand design can have some of that. Right?

Like Style guide.

Style guide. Exactly. And you're building out a series of, like almost like a template world. Right?

Yeah. Where you can build a template, here's what our ads look like, and this is how we wanna kind of operate. And I think that's one way that it could work and you could leverage it. One thing we saw that was kind of popping up that kind of leans to that first thing that I was talking about where people are building their own products is people are building decks now all of a sudden with AI that spit out a a forty page deck.

Then they messaged my brand team, and they're like, hey. Can you style us and make it look like our look and feel? And we sit there, and we're like, no. Like so now we're exploring and playing with ways to like, that initial concept of what I talked about where we can build some tooling.

Parameters.

Parameters exactly that can say, like, hey. Anybody that's building a deck, two things are gonna happen. Either a, give them something that they can use essentially a template, but it's guided via AI to work in the AI tooling that they're working and say, try to stick within these parameters. I think it always needs the designer's eye at the end to kind of refine and make sure things are good, but it gets you eighty percent of the way there.

So that's one way. Or the other thing is of a reality is someone's just not gonna do that. They're gonna make that deck, and they're still gonna message you. And so how can you get your system to take that and convert that deck into using it?

And I think, like, those two paths are something that we've been kind of exploring.

The reason why I say it it's not working well is is more of a broad statement saying we haven't figured it fully out yet. There might be other solutions. The thing that gets me excited from the product design world is the foundational part that does work well out of the gate. Yeah.

Brand design in general, you know, it's just a different a little bit of a different beast.

Yeah. I think Shira has a good question that I think is kind of a nice follow-up to that.

When working with Agencik AI tools, do you ever feel they constrain your creativity? I understand the value of rapid iteration and ideation. I'm curious whether relying on these tools can sometimes reduce the deeper exploration at least in new ways of thinking.

Yes. Great question.

It's it's Yeah. Right on point, and it's Yeah. It's it's ultra accurate. Right?

I think I think it when you're building, it happens so fast. Okay? That's the first thing. And, theoretically, you could just ask it to do something, and it'll do everything for you.

And you can also become lazy as a designer and just kind of be like, yeah. Good enough. Move on. Sometimes things we're doing are good enough, and that's okay, right, depending on the environment.

But I think your ability as a designer to be able to try to push those boundaries traditionally has been, like, messing around. And I had this conversation with one of my designers actually earlier today about this exact thing where in Figma, they were forced to kinda be slow because it was so manual, so they had to kind of think about those things that they were doing versus now, it just kinda spits quickly. So some thoughts around this overall. I think if you're working in this and you're doing kind of a similar model of what I mentioned where you're building a sandbox, you can always work in this way where you can you know, in Figma, you might have done a, this is good enough MVP version.

You might have done the, like, push the boundaries type of version. Keep pushing yourself to do that even when you're working with these agentic tools.

And the nice thing that you can do is you can always build that baseline of what you're trying to do, and you can ask the agentic tool to add a new page in your UI temporarily that duplicates it, and then you start pressing and exploring within that one. Now here's the thing. It's it's not as free forming if you're trying to stay within the boundaries of the design system because you built all these rules to try to keep you Yes. Right?

And so now you're asking it to, like, hey. I wanna play with this and that. And so all of a sudden, you're introducing these custom elements. You can have AI to say, like, let's ignore this for this page.

I wanna explore and kinda do these things. But what I would say for you, Shira, is, like, force yourself to still step back and don't just deliver the minimum that's required.

Be able to put do a version that's pushing the boundaries of it, work within duplicating that same page, and try different things and different versions of it. You can lean I have one designer that leans on the AI to help with brainstorming as well.

Hey. Here's what it is. Do you have other thoughts of what I can do to kind of try to push this a little bit more, explore? And, again, step out of that tool.

Like, we are artists. At the end of the day, at least that's how I like to look at it. And you are looking at your painting, right, on your canvas. The artist that's, like, zoomed in and painting all the details and doesn't take a step back to look at things loses that bigger picture, loses that inspiration.

And so Yeah. Step back from that individual tooling, kind of breathe and look at things and look at inspiration.

I'd like, Pinterest has done a new world for me. My my my brand designer brought brought it into me. I was like a very heavy, old school dribble user for a long time. Pinterest just brought this, like, new crazy cool world of, like, just looking at things that are a lot more exploratory all of a sudden that I could reference and just dive in and keep leveraging their algorithm to, like, see what cool new ways are.

I love it. I think, you know, as a designer, keep training your brain to look at other things and things and work within your own process, Shira, to to not just deliver that one version. You're working so quick now that you could get that version done relatively quickly. Step back, duplicate it, try another one.

I like that. You get whether you're manually pushing the pixel or asking Udenti tool to to push it for you, this the creativity is coming from the creator in a way, and it's either what you know, before it was what you manually painted maybe on the screen, and now it's what you verbally ask or or how you direct this thing that's gonna help you do it or help you spit out some new ideas. So I think there's still room for creativity in there for sure.

Yeah. Yeah.

But, yeah, I like I do feel like.

Oh, sorry. I thought I was just gonna pick up on something that you said. I even think, like, living in, like, the analog world, like, even, like, going to a museum, looking at art. Right?

Yeah. Like, just doing something that takes you out of this, like, digital AI world and just, like, you know, Julia Cameron has that beautiful book, The Artist's Way, which, you know, you have, like, an artist date every week where you take yourself out. No friends, no family. You just take yourself out, and you go do something that's gonna feed your artist.

You know, maybe, like, going to a craft store and buying a bunch of stuff and just, like, you know, making something silly or going to a museum and and just looking at art, but just something that's, like, filling your creative well, which I think will then come through, like, when you're using these tools too. Right?

A hundred percent.

Well, that yeah. Rebecca had a good little question, if that's all that's all right.

Hey, Rebecca. It's my business partner. Hello.

Friend of the friend of the pod. Alan, as your team develops something in house, I think, Rebecca means tooling wise, curious if you know if there are versions of a cursor light out there in the world that's an independent designer can dive in and use. So you obviously built your whole sandbox. Yep. You're independent and freelance. What might be some ways that, yeah, you'd be playing around with the tooling?

Yeah. So, okay. As an example, if you're freelance and you're working with with, with a a client, you wouldn't believe how easy it is to do that first thing that I said. Honestly, I would actually tell you is, like, these things can read screenshots as an example.

That's how I started it. It's like, I took a screenshot of my app. I put it in there. I said, let's rebuild this, but use the design system that I've created Constraints and use these skills as constraints.

And it got pretty far. And I think that's what I would actually say, Rebecca, is, like, lean in and just do it. Just know that because I'm telling you, it's, like, way easier than you think, and you can get something that looks you know this is the other thing. Right?

It's like, depending where you work, like, the exact one to one may matter. Right? If you're in the, like, fast world and care a little bit differently and can get the, like, idea across, this is gonna, like, work tremendously. Right?

If you need that absolute pixel perfectness of it, there's probably gonna be better ways, that can operate for it.

Just my opinion. But I I really think, like, don't don't lean on something else that's out there. I promise you you can make it. Like, I'm talking about a day, and you can Yeah.

Make some main pages of your app or your client's app. Really, imagine your client's face when you go and you have access to it and you recreate it in your own environment that you can start playing and adding things to. It's kind of insane. Right?

That's what you gotta do.

That's cool. That makes me think of just such an unlock as if you were an independent freelancer or contractor and you're working in with early stage startups or something like that.

A way that you could communicate those ideas with those prototypes and just spin that up so quickly now.

Totally. Yeah. Think about it too. Right? Like, if you're if you're interviewing at a start up.

Right? So okay. It's like, we're talking about freelance, like, imagine you can actually go into the admin tool or whatever it is, replicate it, and then do your own and add it on top of it and have a fully working prototype, like, available. You could drop a repo in in the like, you wanna show that you're an AI first designer, drop a repo of their tool that you rebuilt with

Your own value. I'm so glad you said that I was just going to say that. I was like, maybe people can do this in their, like, job applications. I love that.

It's a great idea.

One way to to impress. You know, we've always talked about that over the years of being memorable, standing out, like, what is it about maybe an application or going above and beyond or how you wanna get that attention of a company. This is the this is the modern day.

It used to be you would hand deliver your portfolio with some With a cake. Now rebuild their app and show them, you know, some some feature ideas you had that you could quickly spin up or Yeah.

Yeah. Yeah.

That's right.

Yeah. And it makes them feel so seen too. Like, oh my gosh. Like, you took the time to actually work on our company's brand.

You know, I tell that to designers all the time or people that I'm working with. It's like, you know, hiring managers at the end of the day, they wanna feel special. Like, they want to know Yeah. That you're taking the time and you're taking it seriously, which is tough in this in this modern world where people are putting out hundreds of applications to try to get one interview.

So it's Good. But I think it's it's not about it's not about maximizing the number of applications you send out, but it's maximizing the effectiveness of each one.

Yeah. Yeah.

Doing things like this that really show that you're going above and beyond, and you're utilizing this modern technology that companies know that they need to get. And, like, wow. Like, what a valuable resource to have a designer who knows how to do this that can maybe teach us and show us the way. I think that's huge. Huge. Definitely huge.

Yeah.

Hopefully, we don't offend them.

Yeah. Right.

Well, I mean, we have a design lead But you don't wanna work with them.

If someone if someone were to do that for for your org, Alan, you know, and whether you're hiring or not right this second, someone did what what you kind of suggested there, and and that would catch your attention. Right?

You'd It's an extra value in that. I look. I I sympathize with the designers who are having to apply and not getting getting responsive because there's so many different things. But True. I think, like, in practice, it it really does take out a lot of the manual steps of having to do this that, like you know, even if you did it one page, right, I guarantee you can recreate one page much faster than having to do it manually And taking that extra step to show that that little bit of care. I think I think that connects that makes you smile when you see that res that application and maybe makes that name stick out a little bit more. Right?

Yeah. Yep.

So you would like this you wouldn't get offended like, oh my gosh. Like, why are they trying to redesign my thing? Like, you dig it. Totally. Yeah. Yeah. Good.

There you love that.

You're pissed.

Yeah. Well, we have a minute left to go. Alan, are there any final thoughts you wanna leave people with?

Great question.

Look. I think I think you're gonna continue to see things from everybody, from influencers, from creators, from companies. Companies are big one too. Companies that are releasing these tools that we might have built community and culture around that have their own takes of things.

Everybody's trying to figure out what the right thing is here. The unlock for me and my company was working in the environment that my other peers were working. That's really what helped me and why I ended up in Cloud Code because that was the same place they were. And that's why, like, other tools that even though I was doing stuff and theoretically I could do effectively, if they have to go and download another tool, I've lost them.

Right? If they have to go and and do a different type of translation layer, I've lost them. So I think that's that's that's a big part of it.

I would say just look. No one has the answer. Again, I come out to talk about this because there is an unlock in enterprise that I think has worked for me and my team, and it's not perfect. But it has gotten a lot of the team excited, including our engineering partners who are, you know, I I wish I could quote, but he, you know, he was saying, like, dude, things are moving, like, five times faster now. Wow. And it's not an actual number, but it's not about just the design taking two weeks to do.

It's about think about the entire end to end process now Yeah.

Of reducing meetings, of reducing context, of being able to find the the working file of everybody. When you jump in a meeting and everybody can actually go and see the design that you're talking about and understand and make changes live on the fly, super quickly, get that context in, get that real type of data, in there, the overall cycle gets reduced. And I think that's what companies are gonna start leaning into, and that's where I can see that that, get really exciting for people.

I love that. Yeah. Especially cool too because you guys are remote first company too that you are able to be successful.

I see. Yeah. Yeah.

Mate, this is inspiring. You're really you set a great tone and a great, like, openness to new ways of working, to finding things that work for for yourself, for your team, your org. Yeah. I think it's such a great attitude to have towards just navigating being a designer in in twenty twenty six and beyond.

Thank you. That's all that's all we can do is be there for each other. Right? And Love it.

If you figure out something, I think I encourage you to share as well and try to break it on these same mediums as well.

Yeah. Right on.

Alan, thank you so much for your generosity, everything you taught us today.

It's been so edifying and so wonderful, which is really exciting.

Thank you.

Follow Alan on LinkedIn as well. I know you've been sharing a fair bit on there about your unlocks, about, your process and tools. So I think yeah. Please definitely jump on there and follow Alan so you can see the latest on that too.

Yeah. Yeah. Happy to answer any questions.

Thanks, John. Thank you. Yeah.

Alright. Bye, everyone. Bye.

‍