Kaitlyn Faile on Building AI Tools Without Coding

Kaitlyn Faile shares how she built AI-powered apps using tools like Claude and v0.dev—without writing code. She explores decision fatigue, low-code platforms, and how AI empowers HR pros to build smart tools faster.

https://systemadmininsights.com/

Transcript

00:01 Alex: really excited to be talking to Kaitlyn Faile, who is an HR Solutions Architect on our team, and one of the projects that Kaitlin has been working on is exploring AI app development, which is sometimes referred to as vibe coding. 00:17 Alex: And, uhm, whenever somebody uses that term, I can't tell if they're using it pejoratively or not, but, uhm, you know, I think, I think we have all, all of us who've been involved in this fully understand that what is created requires refinement, right? 00:34 Alex: Uhm, but the, the, the ability to get from, from concept to an MVP is just so easy and so enticing and so much fun. 00:45 Alex: And so today, Kaitlyn, I want to talk to you about what that process has been like, what you've been working on, some takeaways. 00:51 Alex: And specifically from the perspective of somebody who is really curious about how to, how to build their own apps with AI but is maybe, hasn't started yet or is a little hesitant to start, doesn't really know what they're getting themselves into. 01:03 Alex: I thought this would be the perfect opportunity for you to share what you've learned. 01:07 Kaitlyn Faile: Yes. So basically this all started as part of my role, uh, including customer success. So geared towards how do we track what we want to track? 01:17 Kaitlyn Faile: How do we present that to in a way that makes sense to us? How do we get the data in there? 01:21 Kaitlyn Faile: What does it look like? What does it feel like? And I started exploring these no slash low coding options. And just because I was very comfortable with AI as a whole, and I was like, I think I can do this. 01:33 Kaitlyn Faile: And I know how to prompt AI. I know how to make AI get things done for me, reduce my decision fatigue, make me feel and look and sound better. 01:41 Kaitlyn Faile: But particularly, I wanted to use a tool that would let me have a conversation that produced results. And that's really what it felt like. 01:50 Kaitlyn Faile: It was, I knew what I wanted to capture and I could say, please create, because I'm always polite to my AI, 01:56 Alex: even though everybody hates that. 01:58 Kaitlyn Faile: Uh, please create an app that looks like this and has this feel and includes this data, and then it would produce something. 02:05 Kaitlyn Faile: And that was really cool to me to take something that in my mind, I had an image of my mind. 02:11 Kaitlyn Faile: And I just put the words in and then it produced this beautiful looking dashboard that I was then able to say, actually, can we move this here? 02:18 Kaitlyn Faile: Can we change that to this? And then I asked it, how do you receive this data? And it told me what it needed. 02:24 Kaitlyn Faile: And it was like, You can upload it here, or we can just stop and you can test elsewhere and it's not just going to show dummy data. 02:31 Kaitlyn Faile: And I was able to even manipulate it to say, like, include things from my favorite band or my favorite show or things like that. 02:38 Kaitlyn Faile: Just so that the, it looked even more enticing to me in terms of the dummy data that was in there. 02:44 Kaitlyn Faile: So I used a lot of Marvel characters and things like that. Um, so if you know, you know what you're looking at, but, um, really. 02:51 Kaitlyn Faile: The exciting thing for me was being able to say, am I missing anything? Is there anything else that you think I should have in here? 02:57 Kaitlyn Faile: And it gave me feedback even in the prompts there. So I think that was really helpful because I had an idea. 03:05 Kaitlyn Faile: I an idea as the end user of what I wanted it to actually look like and what I needed it to do for me. 03:10 Kaitlyn Faile: And I could get it super, super close with the exception of actually getting that data in there and running all of the things that I needed it to run. 03:18 Alex: That's great. So a couple of things. You mentioned a term that I don't think I have heard before, which is decision fatigue, and I know exactly what you're talking about, but could you say more about that? 03:28 Kaitlyn Faile: Yeah, sure. So my primary for- about stray into AI was I would dump a bunch of meal plan ideas or restrictions in there for what I wanted to eat that week. 03:39 Kaitlyn Faile: And I would say, make me a meal plan that incorporates these things. And then that turned into now go read the salesman's manual. 03:45 Kaitlyn Faile: Ads for the grocery store that I shop at and make a meal plan using only sale 03:49 Alex: ingredients. 03:51 Kaitlyn Faile: And it just did it. And I was like, I don't have to decide what, what. And so that freed up my brain to then go do other things. 03:59 Kaitlyn Faile: So that is my primary use of AI in my personal life. 04:02 Alex: Well, that's great. And I think a lot of people start off that way. It's a great way to start off before trying to think of, uh, business applications, using AI to solve. 04:11 Alex: Decision fatigue in your personal life is something that I did too. My first, uh, GPT that I created was, uh, so, so I lost some weight and I needed to buy new clothes. 04:23 Alex: And so I was buying new clothes, but I don't really have a, uh, a knack for, uh, putting them together in any way that makes sense. 04:31 Alex: And so I uploaded every article of items of clothing that I own, and I created the GPT. Uh, uh, you know, what sort of event am I going to or whatnot? 04:42 Alex: And then it tells me what to wear. It gives me three options that are tailor made to the current weather outside, whatnot. 04:48 Alex: And, you know, it's really like, there's some things that, that. I'm, that I really enjoy doing, um, picking out clothing is not one of them. 04:55 Alex: And so that was my personal project that sort of got me as a starter project, much more comfortable with prompting and how it works. 05:01 Alex: So I encourage anybody to, to, to do that. The meal planning that's, that's. That's a great one too. And, and, and especially, you know, if you're planning for, uh, feeding a whole family, they get very complicated to coordinate all the sales and whatnot. 05:12 Alex: So I love that. Um, so what tools did you use? Was there a particular platform that you found to be helpful? 05:18 Alex: Did you try a couple of different platforms or what was that like? 05:21 Kaitlyn Faile: Tried a couple of different ones. Um, so I really started with Claude because that was, that's my AI of choice, just because I feel like it speaks like me. 05:28 Kaitlyn Faile: It's a little bit more on the creative storytelling. Narrative type side of things. Um, so that's usually where I start. 05:35 Kaitlyn Faile: Um, and then we got into v0.dev. So took what Claude had given me as a starting point and said, here's a starting point of what I'm thinking. 05:44 Kaitlyn Faile: And then boom. Built from there. And v0.dev was a lot more technical, um, but still very conversational in the prompting. 05:51 Kaitlyn Faile: So I was still able to say, can you please do this? Um, can you please make it look like this? 05:56 Kaitlyn Faile: Change this font to this? And I was even able to, even able to give it brand standards and say, make the branding match our website and add in these extra little things that I would, I could probably figure out if I sat down and actually took the time to do the code, but I didn't have to. 06:14 Kaitlyn Faile: So that time was able, I'm able to use that time elsewhere in other ways to be more meaningful. 06:21 Alex: That's great. And were there ever any moments where you found yourself kind of. Going down a rabbit hole or, because this has happened to me where I'm like, this is amazing. 06:32 Alex: And then suddenly I look at the clock and I've spent two hours trying to prompt AI into fixing one tiny, tiny thing. 06:36 Alex: Has that ever happened? And what did you do? 06:38 Kaitlyn Faile: So the main thing was the data for me. So I knew what I wanted it to show me. I didn't know how to give that information to it. 06:45 Kaitlyn Faile: And that's where I really think it's important to have a relationship with a developer you can trust so that you can. 06:51 Kaitlyn Faile: Get it to that 95, 85, 95% completion, and then have someone who actually knows what they're looking at. Look at the rest of it and confirm that it was accurate because AI is language-based, It's necessarily going to be able to pick out every single little thing in the code, but a human can who is trained 07:10 Kaitlyn Faile: in that. So, that's where us knowing what the end goal is and what we actually need the app to do is really, really valuable, because then we can hand that over off and then someone else can kind of pick it up and do the last little techie touches on it and then it's ready to go. 07:24 Alex: Yeah, and I introduced you to Aaron Wagner, who in Ecuador, and Aaron brought home our cost per hire calculator, which is on the ROI page. 07:31 Alex: On the IRD website. And, uh, and so I got to the point, I also use vzeer.dev, and I got to the point where it was pretty much 90% good. 07:40 Alex: And I even deployed it, it was live, but there were certain things that I just didn't have the time or really the understanding to bring it home. 07:50 Alex: And so that's when I went to Upwork and I looked for somebody to work with. And I have just found working with Aaron to be so, so much easier. 07:58 Alex: It's much fun. And she has, uhm, not only fixed the code, so, so there are like systemic problems with the code that AI generated that are just beyond my, my, my understanding and desire to understand. 08:11 Alex: Uhm, and so she fixed that. She made additional suggestions about enhancements to it. She made it mobile responsive in a way that I'm sure would have taken me many hours to do. 08:22 Alex: I'm curious what the conversation with Aaron was like and, and, uhm. Uh, what you took away from that. 08:27 Kaitlyn Faile: Yeah. So Adora, Aaron, um, she and I have very similar mindsets, even though our paths have been very different to get to where we are. 08:34 Kaitlyn Faile: Um, but it was a fantastic conversation where we talked a lot about the. Yeah, a low code, no code kind of status of things, and how much more accessible that is for what we call citizen developers. 08:46 Kaitlyn Faile: So, where everybody has an idea, you can have an idea, but getting it out there and is kind of the first step. 08:53 Kaitlyn Faile: Breaking things is kind of the second step, and then fixing things and actually getting a solid product out there and maintaining it are all, like, the really important kind of things that need to happen. 09:03 Kaitlyn Faile: And she's a great and I mostly discussed that kind of trajectory of, at what point does something need a developer? 09:13 Kaitlyn Faile: At what point do I say, I can't do this anymore, I need some actual help? Because I've broken it beyond what I know how to fix and what AI, I can prompt AI to fix itself. 09:23 Kaitlyn Faile: And then the other piece of that too was, what information can I give you as the developer to produce what I'm looking for? 09:31 Kaitlyn Faile: So. Instead of just, I want an app that tells me my cost per hire, or an ROI calculator that does this. 09:38 Kaitlyn Faile: You have to be able to produce a visual, and that's where the AI tools can really, really be helpful, is you can show them what you're asking for, and then they can actually do the coding side of making it look that way. 09:50 Alex: So, 09:51 Kaitlyn Faile: we talked about specifics of what they need from us on the citizen side. 09:55 Alex: Yeah. 09:56 Kaitlyn Faile: To deliver on the tech 09:57 Alex: side. I love this term citizen developer, and, you know, I'm wondering, what do you think the implications are for HR system administrators? 10:08 Kaitlyn Faile: I think, positively, they would. HR will solve so much of the decision fatigue that comes from doing all the things all the Yeah. 10:17 Kaitlyn Faile: And so system administration, particularly, is reactive. And everything is a fire and everything is the biggest deal. 10:25 Alex: Of anything ever. 10:26 Kaitlyn Faile: And being able to prioritize that in a way that, uhm, gets things done is not something that everyone is able to do. 10:35 Kaitlyn Faile: And I think that's how AI can really support that. You support people who are kind of having to bounce around from all these directions and kind of keep them focused on one thing at a time, prioritize, uhm, and especially with the apps and the way that you can set that up. 10:50 Kaitlyn Faile: You can even include prioritization. Priority levels and kind of indicate within there what needs to be focused on first and what can wait and what is a long-term issue that can be addressed at some point but doesn't need to be done right now. 11:03 Kaitlyn Faile: Uhm, so those are. I prefer to think of it as a way to save time so that you can focus on what's really important 11:09 Alex: and 11:10 Kaitlyn Faile: get done what needs to get done. 11:12 Alex: I agree 100%. So any final thoughts for somebody who is curious? Wants to learn about this stuff? Is maybe having a little anxiety about it? 11:23 Kaitlyn Faile: So, I'm an open book. I'm always happy to connect and share my experiences. I was a high school English teacher before I landed in the world of, Recruiting and TA and rec ops, and now I'm here. 11:35 Kaitlyn Faile: So, I have had an interesting trajectory and lived several lives before this one. And I think all of them have value and point me to where I am today. 11:43 Kaitlyn Faile: So, I am always happy to connect and share and I would also suggest just playing with it. There's really no harm in trying something out. 11:50 Kaitlyn Faile: If you have an idea, see what you can do. And worst case, it just sits in the ether and is there and you can go back to it if you'd like. 12:01 Kaitlyn Faile: But I think most people, if you have an idea and you know what you want, you can produce the results with AI and you can see it and you can get it to a point where you can say, and then make it an actual reality. 12:14 Alex: Thank you so much for your generosity, that generous offer, and for your creativity that you bring to the R&D team, all our customers in our system management insights community. 12:23 Alex: And I am so excited to see. The evolution of this customer success tool and whatever other ideas you have in store, I'm sure it's going to be a lot of fun working together with you. 12:34 Thanks again, Kaitlyn, for talking with us today and I'll talk to you soon.