Screenshots
Tony Naccarato: Aaron comes from a product management background—and in this conversation we talk about how he uses AI to research, design, and build faster than ever before.
Tony: We also dive into one of his projects, Brag Board, which helps workers track their wins and accomplishments over time so they’re ready for performance reviews, promotion conversations, or even job interviews.
If you’ve ever wondered how AI is changing the way people build, this is a great conversation for you.
Tony Welcome, Aaron. Appreciate you taking the time to to meet with me and talk about AI and and your projects that you’re building. yeah, why don’t we just start by you telling us a little bit about who you are and how AI has been working for you and your in your processes.
Aaron Airmet Yeah, thank you, Tony. Thanks for having me on. So I come from a product management background and I have a problem. And that is I have way too many ideas. Way too many ideas of products and businesses that could exist, that should exist, and that need to exist. And for the longest time in my career, I was always dependent upon, well, now I need to go get a technical person, somebody who can actually code, because I can’t code. I’ve tried to code.
I don’t know what it is, but my brain just can’t figure it out. but and I always have to go to somebody who knows how to code and I’d say, Hey, I’m trying to build this thing. Is this even doable? Does the technology exist to do this? Can we start coding this work, collaborate? And then they start to say, Well, are am I gonna get paid? What’s going on here? And so it was just kind of all these weird questions would come up, and it’s just like, no, I just have an idea and I want to test it out or or whatever. but then I was listening a few years ago to this podcast called My First Million. and the
The founder of Replit was on. He was giving an interview. And as he was explaining his product, Replit is one of these vibe coding tools. He was explaining his product. Sean Purry, one of the hosts of My First Million, says, So what you’re telling me is now it’s the time for the idea guy. And I was just like, yes, now is my time. Vive coding has arrived. AI is here. I have all of these ideas. It’s time to go. It’s time to run. So I’ve come from the product space.
Too many ideas and vibe coding and building with AI has started started to enable a lot of different things for me.
Tony I love it. And you know what? I don’t think that you’re alone in that in that thinking. I I’m right there with you and I know there’s many others that that are feeling the same way. I just wanted to maybe like deep dive into one of those projects that that you’re working on.
Pick one and give us kind of a a pitch of of what this project is, like who is it for and what problem are you trying to solve?
Aaron Yeah, absolutely. So I have many, many projects that I’ve been working on. But one of the one of the more prominent ones that is taking a lot of my attention right now, I call it Bragboard. so the idea here is if you are a knowledge worker and especially a remote knowledge worker, sometimes it’s really hard to track your accomplishments over time. And if you work in any size size of business at some point,
Your manager, HR, somebody’s gonna come to you and they’re gonna say, Hey, tell me about your accomplishments for the last quarter, for the last six months, for the last year. Tell me what you’ve been doing for this company. in large companies, it comes in the form of like a quarterly performance review. sometimes they have a form where you have to fill stuff out. For most people and a lot of people that I’ve talked to, that
usually feels like a scramble. I gotta go check my email. I gotta go look through Slack. What meetings have I had? And it’s sort of like, did I have any wins in the last month? Okay, let me do that. And so I wanted to take the idea of how might we help people more track their accomplishments in the moment over time. And then in that moment, in whatever that next high stakes career conversation is, if that’s a performance review or they’re trying to persuade to get a promotion
Or maybe it’s a job interview that they are that they’re trying to look for. Well, now they have this library of wins, learnings, accomplishments, or brags that they could look to over time and pull those out, pull out stories and make that their case for whatever it is is the next step in their career. and so a lot of my early research, turn to chat GPT, perplexity, you know, go look at Reddit channels, whatever, and people are like, yeah, just
like open up a brag document write stuff down in like a Google sheet. Well there’s so many problems with that. One, it’s so unstructured. Two, you forget. Like nobody remembers to open up their their Google Doc thing. Right. I went to Bragboard and I said, how do I make this a repeatable habit for people? So I built in like an email reminder system, a text messaging reminder system.
And just kind of these daily habits of come back and create your brags, enter your brags so that you can have your stories, your narratives when you need them. And then I did something that a Google Doc would not be able to do, which was, okay, you work with people. Like in my day job, I went and worked with some people from marketing. And one of the marketing people said, my goodness, it’s so cool to have a product person who thinks like a marketer. I’m glad that you brought us this idea. And so I said, Okay.
I sent him a unique link and I said, Will you like just put in a quick brag about me on my own personal brag board? And then boom, now his brag gets entered into my thing. And so now it’s like my list of accomplishments that I’m doing, plus all of my different peers that I’ve been asking for them, and I get the reminders, et cetera. and it actually just so happened that in my day job a couple of weeks ago, my boss came to me and said, Hey, there’s people are thinking that there’s some underperformance going on. What have you been doing? And so I go to brag board and I export everything for the last
three or six months. I’m like, well, let me tell you what I’ve been doing. Here’s a here’s a record of everything, plus a dozen or so different brags from my peers so that they see what I’m doing and and then and and it’s just a very persuasive sort of an artifact that you could produce very easily.
Tony That’s awesome. I love that. I would love to dive into just a little bit more of your process of creating this. so like the research, like where did AI kind of first enter the project? or
Or even like where did this project start? Did you I know you’re not necessarily a a designer, but did Figma come into play? Was it just sketching some things out in a notebook or did you start directly in code?
Aaron So I did not start directly in code, neither did I sketch anything out because I am not a designer. I think I maybe wrote out a like a written language user journey at some point. But the idea actually started when I actually went to ChatGPT and I was like, hey, I need a few ideas for things. and and I wasn’t just like give me all of your software ideas, ChatGPT. It wasn’t that I kind of built out this structured prompt of like, I’m looking for ideas that kind of do X.
microhabits, repeatable behaviors, da-da-da-da-da. I don’t want ideas that are this or not ideas like that. So I kind of had to structure that prompt. And then ChatGMT spit out, I don’t know, 20 or 30 different little ideas for me. And then I kind of go through the went through those ideas and I was like, which one of these resonates the most with me? And this sort of micro tracking work accomplishment thing really stood out to me. and then that matched with like, yeah, I’ve been kind of feeling this as a remote worker. It’s hard.
When you’re not visible in your seat and there and you have some peers who are going to the office three, four, five days a week, you know, begin to kind of think, okay, you’re out of sight, out of mind. And so I kind of took Chat GPT’s idea and then morphed it kind of with my own experience and just continue to riff with Chat GPT for a long, long, long time. And it was not necessarily about the solution, but just help me understand the problem more and more and more. and then, you know, you go to a few Reddit channels or a few LinkedIn things.
and you’re like, okay, here’s a conversation in Reddit, Chat GPT. Is this relevant to anything that we’re talking about here? Do other people have this problem? What language are they using? Right. And then you can kind of start to build some of those early, early artifacts. And then at that point, you can turn to Chat GPT and you can say, okay, create like what’s like the minimum viable product for this? What’s the first version of this? What features would it have? What would it not have? And then because I’m a product person, I can kind of sift through that.
And be like, no, this is an MVP. No, this is future. this is absolutely the core experience. Like I take that and mold it and put my own intuition, my own experience, my own spin on it. Right. I think a lot of people just go to ChatGPT and they’re like, yeah, this is perfect. But like, no, take the output, mold it, put in your own judgment, your own taste, etc. And then at some point, when I’ve kind of put my spin on everything, then I say, Okay, ChatGPT, opt write an optimized prompt for replit.
That will take everything that we’re talking about for MVP and let’s go with that. Right. And and often actually before I’ve done that, I usually have checked I riff and collaborate with ChatGPT about what are some different names that I could use for this, like some different brand names. what like what are some design styles I could go with? And I have ChatGPT ask me questions. So it’s like it’ll spit out you know five or seven questions. And I’m like, yeah, I want seriousness, but not stuffy.
Like this is built for individuals, not for HR. Like kind of it’ll ask me brand questions and I and I give it that response. So that becomes really tailored to my ideas and my judgment. I get that prompt from Replit, take or from Chat GPT, put it into Replit. And then now my process has shifted to the AI tool where now I’m like, okay, Replit, here’s this thing. Paste all of this stuff. Let me see what you get on the first try. And then I can go back and forth with Replit agent a few times.
before I get to the product where I really feel like it needs to be.
Tony Awesome. So it sounds like you did spend quite a bit of time in the the research kind of just exploring that with just chat GPT. You didn’t take the first response that it gave you, but you you dug in there. Sounds like you did a lot of research on Reddit and things like that, which is good advice, I think, for anybody that’s kind of exploring this a little bit.
did AI ever give a suggestion or a direction that you hadn’t considered
Aaron yeah, 100%. Like many, many times. Right. And it’s it’s hard to like point to like one example and be like, yeah, that’s that’s the thing. but like for example, when it comes to design, the what is the font? What’s what’s the typeset that you’re gonna that you’re gonna use, right? I don’t know fonts very well. I know enough to know, you know, sans serif versus serif and
I think I took like one design class in college and was like, don’t have more than two different styles on one flyer or something like that. Like, like I there were some rules in the back of my mind, but I am not a Tony Suran expert in this. And so when I went to Replit, I just kind of described Replit like, here’s the feeling that I’m going for, here’s kind of the brand idea that I’m thinking of, here’s kind of the message that I want to communicate. And it picked a font for me. and I thought it was beautiful. I thought it was great. in fact, some of my early alpha users have been like, wow, this.
This is a beautiful UI, and I think part of that is the font that was chosen, right? And I like and intentionally simplistic, intentionally minimal. But the font there is it was actually a really, really beautiful font that I think kinda brings it all together. So that that’s and I ask me what font it is. I couldn’t tell you. I have no idea what it is, but it looks good. And and the alpha users say it looks good too.
Tony even though you’re not a a designer by by trade, like you you know enough, you’ve been able to at least create a prompt that will output a good UI and kind of help you in that process and obviously you’re relying on
feedback or just maybe even some personal preference there to to validate whether it’s good
Aaron So you mentioned a little bit, yes, research. I did do a lot of research, but a lot of research. Let’s put that like in air quotes. Because when you talk to UX designers, many UX designers, I guess not all, but many really love to do user interviews and the usability testing and all those different things. Like I love to do that. Those are fun and amazing things.
But what can happen with some of these tools is you can get into an analysis paralysis type mode. Right. and so the amazing thing about building with AI is you can go through so many different ideas so quickly, and and really get to something that you feel is is super awesome. So it’s like, yes, I did do research. I went to Reddit and I did a couple of different things, but at the end of the day, like you have to move forward, you have to build. You have like
builders are going to move past the thinkers all every time, every day, right? And so UX designers, I know we we like to think a lot, which is good. You need that, but it has to be towards something, it has to be building to something. and you have to get to that point where now I have a prototype. Even if you just feel like it’s a prototype, a clickable prototype that you didn’t have to do in Figma, you didn’t have to connect all the little lines in Figma.
This is actual working code prototype of a web app. Send that out to a couple of people and get some feedback back from that. Like that’s all you need to do. So do your research, but get into a prototype and get feedback as quickly as possible and iterate as quickly as possible. Don’t wait until you’ve done 10 or 12 user research interviews before you feel like you’re in a good spot to move forward.
Tony No, that is great advice. I agree with that. we sometimes just get caught up in well, I you know, have to talk to fifty people before we release this. And and there’s a time and a place for that, right? But I I agree that just you gotta you gotta build, move forward, and I think that’s what I’m trying to showcase with this with this podcast is just
you have an idea that you’re passionate about that you feel fits a need and build it and just get something out there to at least let people look at and start using and then you can start iterating I’m curious about you know what part of this whole process surprised you the most as it relates to to like AI. Like was there anything that was much easier?
than expected or much harder than expected.
Aaron I am consistently surprised. I’ll speak to the tool Replit specifically. it’s the one I’ve used the most. It’s not the only AI coding tool I’ve I’ve used, but it’s the one I’ve used the most. But I am consistently surprised when I give it that prompt after doing my quick research and refining and molding and all those different things with ChatGPT, and I give it that prompt, its first draft of what it spits out, every single time I’m like, whoa, that is so cool.
I’ll give you an example. My wife is putting some plants in the backyard, moving plants around. And she says, Hey, wouldn’t it be cool if we did this or if we did that and we did this? And she’s got this great vision in her mind of a backyard. And I’m just like, I can’t, I can’t see it. Like I just can’t see it. Right. And I’m like, AI could generate stuff for us. there’s a new app idea. Help me envision my backyard, like from a few prompts and give me all the plants that I need and blah blah blah blah blah.
And I’m like, before I actually built out the app, I just built a landing page and I was like, okay, here’s this idea of like AI generation of future vision with my backyard planning, blah, blah, blah, help me plan my backyard. And the landing page it created was beautiful. It created like AI images of like backyards that were beautiful, that looked like real photos, and like there was these little animations everywhere. And I was just like, wow, Replit, like hats off to you. This is awesome.
I haven’t had anybody sign up for the waiting list for that app and there’s other apps out there that already do it better. it was just like I’m blown away by just how Replit specifically has been able to take my prompts and my ideas and create this really awesome visual landing page. Like I did one yesterday for here’s another idea. Why does popcorn only come with butter flavor? Microwave popcorn. It’s sad. Chips have like fifty flavors, microwave popcorn has butter and kettle. That’s it.
So lame. So what do I do? I’m gonna create I l yesterday I created a landing page for a popcorn brand with three flavors of popcorn, right? Barbecue, salt and vinegar, jalapeno, whatever. It’s just a landing page. I just want to see, okay, who’s gonna sign up for some flavors of popcorn? I can connect Stripe, people can pay me pre-orders, and we’ll we’ll validate this, see if people actually want flavored popcorn. And again, Replit just within 10 minutes, I’ve got this landing page, landing page going. It’s got
Backhand integration into Stripe, it’s got databases, I can see everything, and so it’s just it it blows me away every single time.
Tony I love it. what are you building next? I think your your answer would be endless. You
Aaron Yes. As I said, I have too many ideas. Yeah. I have too many ideas, right? So this this Tony is actually one of the I think unintended consequences with these types of tools, especially for product and UX people. We love to tinker. We love to have ideas, right? And so I have known product people who will spend hours tinkering on their AI vibe coded stuff. Here’s the problem, Tony.
Ideas are absolutely worth nothing. And in in fact, vibe coding has just taken the production, the building piece of things and reduced it down to minutes, right? It’s become so convenient and so fast. But building a product is still the easier half of the problem. The other half of the problem is distribution, it’s marketing, it’s sales, it’s convincing somebody to pull out their card and give you money.
For your product, you have to get their attention. You have to like, it is so hard, right? I haven’t cracked that nut yet. And so what do I what do I have? Is I have a graveyard of ideas where I’m like, well, let me just spin up a landing page and and see if anybody signs up for the wait list. No, it’s so dumb, right? Because even just that, well, now I got to do some paid ads on Google or LinkedIn or something and just drive traffic to that. And so it’s like, I have not yet figured out that half of the equation. I can build product.
all day long. It’s fantastic and I love it. And I can think through new features. And Chat GPT is always going to give me new feature ideas. It’s great, right? But getting people’s attention, convincing them that you have a valuable solution for their problem that they’re willing to pay for, like that is the much harder half of the game of building a product.
Tony Totally agree. I’m in the same boat as you with that. It’s ideas come freely to me and I can build them. but getting people to actually use them and marketing them is is the challenging part.
Aaron One one rule that I try to follow with that, this comes from Noah Kagan, who wrote a book called The Millia Million Dollar Weekend or something like that. And he’s kind of this entrepreneurial business guy. but he talks about, he calls it the law of 100. and so I’ve kind of interpreted that as okay, if I build a product like bragboard.work, then I need to tr attempt to have 100 conversations about my product with potential customers.
Or at least reach out to one hundred people about this product. And if if the market and if they’re not telling me anything or if they are telling me something or whatever, and I can iterate from that, but if I if I attempt to have a hundred conversations and I get nothing back, then I’m thinking this idea probably is not there’s not a market for this idea, at least to how it’s currently packaged and or messaged.
Tony Yeah, great advice. and great book too. last question and let’s focus on your bragboard. If you had to start that over tomorrow, what would you do differently?
Aaron that’s a tough one.
If I had to start it over tomorrow, I probably would have spent less time building out features that I thought were valuable. And I would have spent more time. Once I actually had a workable prototype, I would have spent more time trying to get feedback from early alpha and beta users. I got a few Alpha users on there. They gave me some good feedback. but
Again, I probably would have focused more on the distribution side of the house than on building more features, building more features, building more features. because I have some features in there. Like I’m the only brain behind the product. I have features in there that are useless because they were just my brain, right? and and like I spent time and money building those when maybe I could have spent time and money trying to do more validation conversations.
For all the the dumping I did on research earlier, right? Like validation conversations and distribution tactics techniques, like that’s where I would spend more time and money now.
Tony Great advice. Great advice. Well, Aaron, thank you so much. definitely appreciate your insights. sharing your experiences here with the products that you’re building.
Aaron Absolutely. Thank you so much for anybody who’s a remote knowledge worker who needs to track their accomplishments. Brag Board.work. And good luck, everybody. Go design it, go build it. And it’s the time of the idea person. So go forward, feel enabled, don’t stop, get out of your own head, get out of your own way and just start building stuff because you just never know where it’s gonna lead.
Tony Great advice. Thank you. Love it!