Episode 26

full
Published on:

18th Nov 2025

Sarah Medilo: Building AI Agents and Balancing Family in the Philippines - Ep 26

Join us for an inspiring episode of Builder Stories featuring Sarah Medilo, a former CEO turned AI enthusiast. Sarah's journey from leading the top HubSpot partner in the Philippines to embracing AI technology is a testament to innovation and personal growth. Her story is not just about technology but about finding balance and purpose in life.

In this episode, Sarah shares how she transitioned from a successful career to focus on family, particularly her mother battling Alzheimer's. Her return to the tech world was marked by a newfound curiosity about AI, leading her to create practical solutions like the "go bag prepper" agent for emergency preparedness. "I realized that hey, you can build agents too," she reflects, highlighting the accessibility of AI for everyone. Sarah's approach to building AI agents is rooted in her experience with workflows and process optimization, demonstrating how technology can enhance productivity and foster deeper human connections. Her story encourages builders to embrace technology without fear and explore the potential of AI to enrich our lives. Whether you're a seasoned developer or a curious newcomer, Sarah's journey offers valuable insights and inspiration.

Learn more and connect with Sarah Medilo:

Subscribe for more Builder Stories and visit agent.ai to start building your own agents.

Transcript

I felt like, hey, this is something that you need someone who's a developer who has a agent building mind. But then I realized, right after the workshop, I was like, I can do that. That's simple. I mean, way to go there mesh for making it drag and drop. It can start with simple passion projects like the Go Bag preppers, and then turn into something that will really give them back the time that they need to get things done more and to be present to people more.

And in this episode, we meet Sarah Medilo, a RevOps professional from the Philippines who turned a late night workshop and a moment of inspiration to an agent that just won "Most Useful Agent" at a recent builder workshop with her Go Bag Prepper agent. With 20 years in tech and a heart for service, Sarah's redefining what it means to build technology for good. She spent time at Apple and Adobe, building for the giants, and now she's leading RevOps at mission-driven non-profit organizations. Sarah brings a unique perspective on how AI agents can save time while amplifying the human element that matters most in business growth: relationship building. And with that, let's get to the show.

**Kyle James:** Sarah, welcome to another episode of Prompted Builder Stories with me. I am excited to hear your story. So today on the pod, we've got Sarah Medilo, who went through our workshop course to become an agent builder. And just the experience through it, Matthew reached out to me and said, "Kyle, you've got to talk to Sarah. She's got such an interesting story." And the other exciting part of this conversation, this might be the first international conversation we're having. So you're in the Philippines. We were just joking before the call. It's 8:00 p.m. there. It's 8:00 a.m. here, but yet it's raining in both places. So the world is as much of a big place it is, it's still a small place. So I'm going to be quiet now because Sarah, I want to, I want you to introduce yourself and tell a little bit about, you know, your journey, which is absolutely fascinating. You're telling me about, you know, your journey getting into AI, kind of your time around the HubSpot ecosystem, and and the break and then coming back and realizing, oh my god, I got to figure out this AI stuff. So, please, like, where do you want to start?

**Sarah Medilo:** First, I I have to greet you guys in our local language, Mabuhay. That's welcome in the Philippines. Mabuhay to everyone here listening. I love that we are a global audience now. Um, we have we have guests now from the other side of the world. We're exactly on the other side of the world.

er in the Philippines. And in:

And that was:

But in June of this year, I had a conversation with a founder, well, an owner of a VC, a venture capitalist uh company, who was doing work with mission-driven organizations. Oh, cool. And I felt like, hey, this is the best fit for me. I know I'm I'm in a stable um role in a bank. It's hybrid, and I'm making impact, but this is very interesting because it's very close to my heart.

Kyle, I actually, um, I actually am a uh board member for several nonprofit uh organizations in the Philippines. I have a I have an an NGO for peace building called Teach Peace, Build Peace Movement. I have a photography uh NGO who helps kids in like rural areas, like up in the mountains or down down south. Um, we teach them photography. We actually exhibit their work in their communities. So, you know, it's a different perspective when you're looking at kids taking pictures and showing their community. And then I'm I'm also part of uh, well, I'm the movement integrator of Strengths Philippines, which is the NGO that manages over 300 Gallup Strengths coaches, all certified here in the Philippines. So, about 300 of us, um, and we just recently celebrated our 10 10 years uh, with our Empower Power Con conference. And so when I heard about, Wow, that's stuff. You're doing a lot of stuff. I'm a big activator right here. So, um, when when when um, I got into a conversation with the owner of, one of the founders of Foundry for Good, which is the VC I'm working for, and he spoke to me about the work that he does with um, with mission-driven organizations, I really was was attracted to it.

So, I made the the the leap to to actually work for Foundry for Good, where where I actually handle four four of our companies, um, and help them with their RevOps motions. And honestly, you know, I was away from the HubSpot world and all the the technology innovations that are going on for about nine months. You were in the real world, not in this tech bubble. Yeah. Yes. In in those nine months, when I came back, I was like, what's going on? There's so many updates. And I I I had FOMO. Like, I felt like I missed out. Like, a lot of things I had, I felt like I had to catch on up on. I was listening to all the updates of George B. Thomas and Katie Hawkins and all of them, just making sure that I I understood what the product updates were there were happening were.

And then I realized also, apart from the changes that HubSpot was doing on the product itself, there were a lot of changes also that were actually mixed with the development of AI and how AI is really like changing the way we do work. You know, there on the HubSpot side, you know all the updates on AI prospecting agents, customer agents, and all of that. Um, breeze into HubSpot. Um, I heard, I knew about Agent AI because before, right before I I took a career break, I was actually in Boston and I saw Dharmesh and Yamini and Brian, and I caught Dharmesh right after his his keynote and he had just announced the age of AI. And I knew Agent AI was coming up.

So, I was excited about it at that time. I wanted to play around with it, and then my personal story happened. Um, and I had to take a break. And I knew Agent AI was there. So, I actually, like 2 months in, um, with the VC, checked out Agent AI again. And um, I was looking at all the agents that were available in the space. And I actually started enabling our sales teams with use, the use of some of the agents there, like company research agent. There was a LinkedIn profile summary agent, and then the objection howl agent. I mean, there were so many agents there. So, um, I actually started enabling the teams with those agents, just giving them options of which agents they wanted to use, wanted to like bring in as part of their process when they're trying to work with the humans that they were that they were engaging.

And it was only just a few weeks ago during the October um workshop with Matthew and Mike that I realized that you can build agents too. So, you know, their workshop was uh midnight my time. So I stayed up just to make sure I was like listening in there. And um, I I tried to figure out, hey, this is something that I can actually do. I can actually build agents. I mean, I know that um, that that's possible for people who are developers. You know, that was my, that was my bias. That was my bias at that time. Um, I felt like, hey, this is something that you need someone who's a developer who has a who has an agent building mind. Sure. Um, to do. But then I realized, right after the workshop, I was like, I can do that. That's simple. I mean, way to go Dharmesh for making it drag and drop, you know? Yeah, it it's it's literally, if you can build a workflow, you could build an agent. Exactly. So I I was like, okay, I mean, I build workflows for what I do all the time for all the different processes that we need and making sure that we save time for people by build, by automating some of the the admin work that they were doing.

But this is different. I mean, this is with the use of the current LLMs and all of that. And I never realized that the way Agent AI was built was it wasn't just banked on one LLM. I, you know, in my mind, it was just simple one. And then I realized, oh my gosh, they have all the open AI, perplexity, they have, they have Claude, they have Gemini, they even have DeepSync. I was like, it's all in there. So, right after the workshop, they actually, there there's a slight contest. And I wasn't really, I wasn't really like intending to join, to be very honest, open and honest. I wasn't intending to join. Why not? Why not? You didn't think you could or you just... Not just that, but I felt like, hey, maybe, maybe I I, you know, I I still need time to like perfect an agent to make it public. Uh, but when I woke up that morning, my head was like in in this mode of, I want to play with that tool. Yeah, you have to give it to me, Kyle. You know, it ended at like 1:30, so I wanted to sleep. Right? Your brain had just absorbed all of that wonderfulness and you needed to let it rest a little bit. Yes. And then when I woke up in the morning, um, I had, I didn't even have my coffee yet, but I was thinking about it. I was thinking about it. I was like, what can I build? What can I build? And, you know, I was thinking what can I build without even thinking of the contest. And then, um, had my coffee, was in front of my computer and uh, I started thinking about what should I build.

So, at first, my first idea was actually to build a an agent that would pull, um, which, you know, this was the time where when here in the Philippines, we had several earthquakes happening. So we had a big one that happened, 6.4 that happened in Cebu, which is in the middle part of the Philippines. There was a lot of devastation that happened there. And then we also had a recent one, like two days prior to the the workshop. There was one that happened in Baguio in La Union. So that was in the northern part of the Philippines in Luzon, our biggest, um, one of our biggest uh islands. So, you know, everyone was talking about earthquakes.

So I was thinking, maybe I should build like a an agent who would pull information from PAGASA, our weather station, and just give updates on on where the earthquakes are. So that's my initial idea. So, I'm an activator, so I just went on trying to build it, you know? I had I had the idea in my head. I just wanted to try and build it. And then I I was actually getting challenged, cuz I wasn't sure like, how how should I build this?

And so, I, you know, this is still in the morning, maybe like 8:00 a.m. Like by 9:00 a.m. I had given up on trying to build the the the earthquake agent. And then I was just walking around the house and I said, "Why are you making it so difficult?" In my head, you know, I said, "Why are you making this difficult, Sarah? What's the one thing you really need right now?"

And then my husband messaged me on Messenger reminding me to make sure I had packed all the right, like in Ziplocs, clothes for my kids and I. And I was like, I saw his message. I was like, "This is what I need. I need a go-bag prepper." I need, I need um an agent that will help me list down all the things that I need to prepare for my husband, who's an adult, my two adult girls, my son who's 7 years old, my mother who's a senior, you know? And in back of my head, I should include like the the babies in our family, like in our cuz I was meaning to share it to our family, you know?

So I was like, "Okay." I took out my notebook. And this, I really appreciate how like Mike um in in our in our workshop was driving in. It's it's best that you plan it well. So I remembered that. And so I took out my notebook, wrote down like ages it should be. And then I said, "Okay, what else do I need?" I need to figure out like what uh categories I need. So, like, list it down. Food, water, supplies. Food, water, you need medicine, you need equipment, you know? So I listed that down. So, when I had the plan in my notebook, I sat down again on my computer and built it, and I was done in like what? Maybe 15, 20 minutes? Measure twice, cut once, right? That's what the carpenters always say. So you had figured out exactly what it was going to be and done the planning. So when you went to go build, you knew exactly what you wanted to do. Yes. And you kind of ideated a little bit. Yes, and it's a very simple app. I mean, granted, when I first um invoked the the LLM, I had issues with the output that it was giving me. So I had to I had to change my prompts. I had to put more guard rails. So GPTs typically ask you more questions to want things. So I had to put guard rails so it wouldn't ask me more questions. And um, maybe I I I think I ended it twice only, my prompt. Yeah. And then once I was done, I was happy. And I was like, "Oh, now I think I'm ready to like submit this to the contest." Yeah, and that's when I submitted it. So, it was probably like early morning your time when I submitted it, but I was done with it before lunch. And I had lunch and then in the afternoon I submitted it. Sometime after lunch I submitted it. Yeah.

**Kyle James:** Wow. So, that's crazy. So, a good night's sleep and then a couple of hours and and you'd built your first agent, never having built anything software related before. But but like you said, it's, but I'm sure you've built many workflows before, right?

**Sarah Medilo:** Workflows, yes, because that's something that I really do when I think about like helping people automate certain processes, you know? Uh-huh. Understanding. So, when I think about it, it's like, it's very related to the RevOps work that I was doing because the experience of, the experience of building the agent really was understanding first the process, what my goal was, right? So, um, and typically when I, when I build workflows for my teams, I have many questions about what is the current process? What do you, where do you want to bring it? Yep. What are the other things that that that probably impact what we're going to do, right? So I try to understand all of that.

So, when initially, you know, I I mean, I'll be honest, my initial experience was, let me try and play with this, right? When I was trying to build that earthquake app. And now when I, when I think about it, it's really funny because they can pull that on Facebook. Why would they want that on an agent? It's not. I I felt, I felt like maybe I wasn't really thinking about the problem, what I wanted to solve really at that time. But when I, you know, I stopped and I thought about, what do I really need right now? And the go-bag prepper was what I really needed. And because I needed it, maybe that helped also in my motivation to get it done.

**Kyle James:** First of all, we should definitely demo this, but before we demo it, I want to take a step back because you said a couple of really interesting things I don't think I'd put together until now because you were kind of a power user of of of these things, right? Like, not only were you using them, but you were like finding the good ones and and introducing them to your teams to help them solve problems, too. Which is I think a really interesting angle. But I also, I mean, I told you this before we started, like, I personally have a mission and believe every person can become a builder. Um, but you start, you know, you start using stuff and playing with stuff and trying stuff. And then you get to the point where there's some agent you're like, "Oh, this is great, but I wish it did this." And then you kind of build it or or, you know, this, it's like, I need, I need to do something. And, you know, a go-bag prepper has nothing to do with RevOps, right? But you had this problem that you needed to solve completely unrelated, and you wanted to play with it a little bit. And but I'm I guarantee you, now that you've done that, that play has probably enabled you to do some other stuff that you in your role and in your job that you would have probably never thought about before, right?

**Sarah Medilo:** Yes. I mean, you know, to be honest, the the the Go Bag Prepper is so basic. In fact, um, for for my, for what I want to be able to do, I think I'm okay with that already. But I'm sure like other people could like spin this around and make it even more relevant to what's happening to them, customize it for them. And that's possible um, with with Agent AI. Um, but on on my role, specifically, lately I've been building some scorecards for my for our teams, um, because they they work with different types of um organizations, all mission-driven, but they help them with things like SEO, or maybe things like AI visibility even, um, or um matching gifts. These are all like um what what what um the different companies we serve do.

So I started building some scorecards uh for our internal teams. I'd say I'm about like 60% in on them. And the only reason why it's not taking me as quick as I was doing it with Go Bag Prepper, what is because this time I'm very intentional about the the things that I want uh the agent to be able to drive. So I'm intending for this agent to be like an internal resource for the sales team. Typically, when they work on scorecards like this, we do have a scorecard metric right now. It takes them about 3 to 4 hours to work on one scorecard for one company. So I'm working very closely with the teams right now to try to understand exactly what they want the scorecard to build. So it would be aligned to their expectations of what numbers they want to be able to show the people that they work with.

So, the intention is build this and allow the agent to run it for them, the 3 hours down to maybe 5 minutes, 2 minutes. Oh wow. And and I think for that, the big, I think, my, the the big thing that will, that it will change for our teams is they they will actually have more time to speak with humans and be more intentional with their conversations with humans, rather than going back, research, running the scorecards, tracking them. I mean, it's not manual because, you know, we have technology, but the research part and putting it all together and bringing together the the meat of what they want the the scorecard to be of value to people they speak to. I think that's it's what I really want to be able to help them with.

**Kyle James:** I I love that. And and I'm hearing, I'm having a lot of conversations around that recently, that agentic AI done right is more, is it's, it goes back and does the research so that you can spend more time on that verification with a real person, right? The relationship building, because AI doesn't build relationships, humans build relationships. And and and once we've got the busy work or or like the research or whatever it might be, done and brought to us so we can kind of consume it and then have that meaningful conversation. But we're not spending hours trying to like put all that together anymore.

**Sarah Medilo:** Put it together. Exactly. Exactly. And I think, I mean, you know, before I left uh, before I took a career break, that was the fear of people about AI, that AI was going to take jobs. But now that I'm seeing how AI is actually transforming and how it's developing, I really feel like we can use this for good. We can use this to, in a sense, allow us to extend our time for meaningful things. Because if you can get an agent to do, like you said, all the research, all the data crunching, all the computations that have to happen, and with the right guard rails, of course, and be able to look at the output and say, "Okay, this is something I can use now," then it's really just a an addition, an addition to us, right? It helps us become more relevant and meaningful in our conversations. Yeah. It allows us to, for my sales teams, for our sales teams, that's what I'm hoping, that they get more time to be in front of humans because that's one thing that the agent can't do. Like an AI can't do, right? It's really being present. And I think this development of AI could really allow us to give us back more time to be able to do the more important stuff.

**Kyle James:** I think it will. I think it will. And I think that's the right, that's when it's done right, I believe. It's like human in the loop, that's the term I I like to use that's correct. But really it's like humans over the top of it, but if the AI can make sure that your stuff gets into a CRM, because the sales reps, we all know sales reps hate doing that. They don't they don't want to do it. So if the AI can do it, then you're happy as a RevOps person because it's getting done. The salesperson's happy because it's getting done and they don't have to do it. And everybody can win with the AI kind of stitching this stuff together. And it's not, it's not replacing people, it's making everybody more efficient so more stuff can get done.

**Sarah Medilo:** Exactly. Exactly. Yeah. And yeah, it's really, it's really a a great tool for people to learn also from their experiences, especially with AI being able to pull things like conversational intelligence, you know, from your CRM. And now, you know, you can build a smart property that tracks uh customer champions. And, you know, coming into coaching calls, you you can, I've taught our sales teams to look at, then talk to your teams about who the champion is and how you're going to build the champion. This is the, this is the property that will tell you that. Yeah. And, you know, these sales leaders, they don't have time to listen to call transcripts, like hours and hours. No. Yeah. AI can do that, right, for you. So, it's amazing.

**Kyle James:** Yeah, summarize it and give you the stuff that you need to coach them on. Like, "Oh, they're they're weak in these three points." And a lot of times, I imagine a manager kind of knows intuitively, "Yes, I need to work with them more on listening or or objection handling or whatever it might be." But, you know, the data, the data that comes out of the AI, the the LLMs can kind of verify that. So, like, you, then you're like, you double down, and and the person, the the person in the role can like, "All right, the manager's going to help me with this. The thing's telling me I need to do this, and it's, it's not, it's not, it's not biased, right? It's unobjective because it's just looking at the data and bringing out stuff."

**Sarah Medilo:** Yeah, that's why I love it. It can really make us see things more comprehensively, I would say. You know, things that we never thought, we never saw before because we never had the time.

**Kyle James:** All right. You you want to you want to demo this thing? Cuz like, I know that you're like, Sarah's like, Sarah's telling me before, it's like, "Well, it's just a basic thing." I said, "I I understand it's a basic thing, but it just kind of shows that anybody can get in here and with a little time build something basic. And the like, what's an analogy here? The very first video game built was what? Like Pong or something? Now look at video games now. You know, none of those things don't exist if Pong didn't exist, if Tetris didn't exist, if these early age video games, you know, look, look at look what Mario is now compared to what the very first Mario version was." So, yes, this thing might be basic, and and Sarah, whether you want to take it on to all the things that it could be or not, like maybe someone sees this and like, "Oh, I really want to build this thing that looks at different types of weather and and and and gives recommendations around that or different locations and and, you know, like it it's just a spark point." And I think that's what's really cool is you just went in and and had a problem and played around and got something that did what you needed it to do. And then where does it go from there? Maybe it just all goes into like, "All right, I'm going to go back and build RevOps tools from what I learned and played with here," or, "I want to take this to the next level." There's not a wrong path there. How many steps is your agent?

**Sarah Medilo:** Three steps in the workflow. Yeah, yeah. But two steps for the user. So, the, all the user has to do is decide what age the people they're creating the go-bag prepper for. So you can decide, you can, you can choose from a baby to a child to an adult to a senior person. So their needs are very different depending on age. So, that's the reason why I built it. And, you know, when I was building it, I was actually thinking about the people in my household. So, I have two adult, young adults. Um, I have uh my husband, I have my mom, who's a senior, and a son who was, who's 7 years old. And then I was also thinking about the people in my family who might need it. So I was thinking about the babies in my family.

But when when I was building the app, Kyle, I also had to think about like what categories am I going to drop? Am I going to share? So, I I've provided the agent instructions to list them down according to food, medicine, drinks, clothing, um equipment, important documents, um so I had the agent write that. So, You got six categories here. Nice. Yes.

So, for different ages, it will be also, you know, different, a different list. So I I I I expected that. And in the beginning, when I wrote my prompt, I just wrote a simple prompt and I didn't put enough guard rails. So at the end of the output, the AI was asking me, "Would you like me to create more?" So I was like, "Oops, this shouldn't work, right?" So, so I I added a guard rail telling the, telling the agent not to ask extra questions. And then I also, so in the beginning, the documents were very little. I mean, they they didn't even write the passport. They didn't even write, they wrote the birth certificate. So I wanted it to be more more complete. So I made sure you ensure that you have the passport and other health insurance documents and all of that. So I added that to the prompt.

So, this one, the recent photo for child and caregiver, I actually added that because I wanted to make sure that there was that information available in the imported documents because it's a child, right? And at the end of the day, I was able to let the agent come up with the list that I wanted. So, it's it's not, it's not as simple as write a prompt. You do have to put some guard rails in so that you ensure that the LLM comes back with the right output that you need.

**Kyle James:** But we call this in the software world, we call this a minimum viable product, an MVP, right? You get something started, you get it to work, and then you can build upon it. But I think it's a great example of of just, you know, like you said, you, once you kind of knew what you wanted to build, you built this in in no time at all. And and you got it to work and and let it, you got it set up where I could go run it right here in front of all these people watching this, which is, I think a really cool, incredible, you know, incredible outcome here.

**Sarah Medilo:** It's very simple. Yeah. I'm sure my, I'm sure my this would be very helpful for my my fellow Filipinos who are undergoing a lot of calamities right now. You know, that there was another uh volcano that erupted just this Sunday here in the country. Just last week, there was another volcano also, Mount Taal this Sunday and then last week was Mount Kanlaon. A lot of things happening in in our area very close to the Pacific Ring of Fire. So I think this would be helpful for them. I realized also this would be helpful for people in your part of the world. Um, I know, I know a lot of typhoon or well, tornado alerts happening also in in the US. So I'm sure it will be helpful as well.

**Kyle James:** Yeah, I mean, I'm in South Carolina, like we were talking before the call, and it's been a really quiet hurricane season for us in the Atlantic. But I think that's a perfect example of of if you wanted to build this out or somebody else wanted to pick up this mantle and run with it, what kind of emergency are you dealing with, right? Because I imagine what you need to pack and prepare for a fire is very different than a flood, which is very different than a blizzard or a tornado or or an earthquake. And and some of these, you know, depending on location, might introduce multiple things, right? Like, like a a earthquake might also introduce a a tsunami or something. So you've got to like think about those kind of things. Um, which which is a whole other level that gets more into the granulars of what it comes out with. But yeah, I mean, like I said, this is a a minimum viable product, an MVP that has a lot of places you could go, but just to show, like, you were able to put this together in in a few hours.

**Sarah Medilo:** Yeah. Well, I'll be honest with you, Kyle, I'm sure Matt and Mike will run another workshop in November. You know what my, my ask would be? Maybe the November workshop-ers can like build on this agent, add those uh particular use cases. Because I know you can, I I'll turn on this, this, this uh agent so that other people can can build on my build. There you go. Call to action. amazing isn't it, right?

**Kyle James:** Call to action. Yes, if if somebody is struggling with an idea, Mike and Matt will make sure you guys have access to this one. They could use it, they could build upon it, they could look at the innards, you know, see what what how it's working, and if they want to add additional steps and and better refine it. Yeah, it's a great thing. And, you know, maybe maybe I'm sure what they'll counter with there is like, "Hey, do you have anybody on your team or or or that you want to send and and want to build it out more, too?" So, you know, fun challenges all around.

**Sarah Medilo:** Well, well, one thing that I learned from the workshop is you can actually invoke other agents within Agent AI. So, um, I think that's amazing. So I I would imagine, like the next version of the Go Bag Prepper would be something like an AI who uh someone who built an AI who's going to invoke the Go Bag Prepper to do something bigger. Yeah. Right? So, it's amazing.

**Kyle James:** Kind of bring this back. You you talked a little bit about you're working on some kind of scorecards and stuff. How do you see this, you know, I guess getting back to the RevOps, which is kind of, you know, your superpower, how do you see this playing out over the next 18 months, 2 years? You know, how how much more of this stuff do you see coming in? And how do you see it changing the work that you do fundamentally, now that it's kind of opened your eyes to what's possible?

**Sarah Medilo:** Well, to be honest, I mean, the scorecards, I think, are just the beginning. These are things that we had. We had a a process for it. We have a way to manage this information. I thought coming in in June, when I saw the scorecards being rolled out by to by the sales teams, using, they're using this in their process, I felt that my my what I needed to do was actually build an, I need to build an app, like that I could put into our website and allow it to like compute all these things, or even like, maybe even like build workflows that process that some way somehow, come up with the output that the teams wanted.

But right after the workshop, I realized, hey, I can build an an agent for this. I don't have to do the heavy lifting for this. You could have a workflow that does part of the stuff that maybe doesn't require LLMs. But a simpler workflow. I don't have to build a complicated workflow anymore because I could I could really see my AI agents being able to uh compact the information and summarize it for the team. And then maybe use the workflow to like deliver it to to the clients, right? Um, and set up the meeting, you know?

Um, so there are many use cases that I I see our teams can use internally, but I'm also excited to be able to, in these conversations with the sales teams, be able to understand more of what they hear from our clients as well, because I'd love to be able to build like agents for clients so that they could make better decisions on things. Wow. Yeah, yeah. Right? So, that's that's where my head is at, and I think over the next, we'll see more and more teams being able to use agents in their day-to-day. It can start with simple passion projects like the Go Bag Prepper, and then turn into something that would really give them back the time that they need to get things done more and to be present to people more.

**Kyle James:** There you go. It's all about how do we spend more time with each other, building deeper relationships. Yes. As we kind of wrap up here, you know, you obviously threw out there, "Hey, anybody in the community that wants to pick up this thing and kind of take it to the next level," but how else can the community help you? Like, what other, what's the best way to connect with you? What sort of help, support, customers, whatever, are you kind of looking for? And and kind of share that with people.

**Sarah Medilo:** Well, people can always look, look me up on LinkedIn. Uh, so my name is Sarah Medilo, S A R A H M E D I L O. I love um getting connected to people, sharing ideas, which is the reason why the community, the Agent AI community is also now a new place for me to be able to get ideas, share ideas, um, and be able to commune more with people who are like me, curious and um want to learn more about this new resource that we have um in in our belt. So, just hit me up on LinkedIn. I'd love to have conversations around RevOps, if you're RevOps practitioners. HubSpot, if you love HubSpot like me. Um, and also your experience with AI and really how it's transforming the way people work and how we can all thrive together in this new age. I mean, being a Gallup certified strengths coach myself, Kyle, change management is something very close to me. So, I think it's important that we connect and and and and have conversations about how um we can make the best use of this new AI that is in our workspace, and also how we can um be able to think of newer ways to be able to make deeper relationships, deeper connections with people that we work with and people that we serve.

**Kyle James:** I love it. I think that's a great takeaway for everybody. To use the AI to allow you to have deeper time and be able to build deeper relationships with people because the busy work's getting handled. Yeah. And and Sarah, I I am absolutely sure that your mom must be very proud of you and all the stuff that you're doing and and and the way that you're finding time to build deeper relationships with people. I'll kind of leave you with this as kind of closing. Any final words of wisdom or things that you want to leave with the audience with or or kind of a takeaway or or maybe something that I didn't ask that you just wanted to share with everybody?

**Sarah Medilo:** I think if there's a message I was going to say to people, don't be afraid of technology. AI is just another technology that we we have now. I grew up in a world where we had black rotary dial phones and we had to, yeah, we had cassette tapes for our songs. And now, like, you know, with technology, you can just choose whatever you want to watch, you want to hear. Facebook, um, came out when I was, when I was at Apple, actually, I was part of the the pilot team that worked in Apple with Facebook. Wow. So, I was with the education team. So, I had Facebook earlier than anyone else here in the Philippines. So, that was new, also. People were afraid of it as well. Uh, even up to now, there's like a back and forth there with uh what's happening with Meta.

But these are all technology, and technologies are meant to make our lives more, better, to to allow us to do more things, to allow us to have more time. So, your mindset with these technologies really just has to be that you use it for good. And be intentional with the way that you use it. And as you use it, be responsible also of understanding how the technology works so that you can ensure that it's supporting you with the intention that the good intention that you have, you've meant for it. So, Yeah. I love it. It's exciting times for us right now. I can't imagine what the world would be for my 7-year-old. He's probably going to see like jets flying, or we're probably going to go to Mars, right?

**Kyle James:** I know. I'm right there with you. My youngest is 7 too, and just like, I don't even know what what they're going to look like, what it's going to look like by the time that they're our age. But But I think it's a great message, everybody. Go out there, use AI for good, build stronger relationships, and and yeah, keep growing and keep building. And until next time, do those things, and we'll see you next time on Prompted Builder Stories. Take care, everybody.

Listen for free

Show artwork for Prompted: Builder Stories

About the Podcast

Prompted: Builder Stories
Builder Stories is an official podcast of Agent.ai, where we spotlight the creators behind the agents. Each episode shares the journey of a different builderm, many of whom aren't traditional developers, showing how people from all backgrounds are using AI to solve problems, launch tools, and build their way into the future. If you're curious about what’s possible with AI agents, this is the place to get inspired.

About your host

Profile picture for Matthew Stein

Matthew Stein