AI Automation Without the Hype. Real Nearshore Workflows That Scale

Foreign.

Welcome to another episode of the Nearshore Cafe podcast. I'm your host, Brian Sampson. And if you're interested in how AI and nearshoring are merging, this is going to be a really good episode for you. Learning about operational upside, workflows, efficiencies, ways to improve. It's going to be a really educational show. Before I do that, let me thank our sponsor. That's Plug Technologies, plugg Tech. Great way to connect talent from all over Latin America with US Companies. Welcome Marvin Martinez, CEO and founder of Bandsaw AI.

Hey, thank you, Brian, for having me. Your show, it's a pleasure.

So I want to be transparent here. I was so impressed working with you, Marvin, for the last year as a client that I've decided to invest recently into Bandsaw. I think it has this really exciting arbitrage opportunity of all that we already know about Nearshore combined with AI. I think about this phrase, double bonus arbitrage. Tell the audience, what is Bansol? What does it do? Types of customers it works with. Tell us. Tell us all about what. What's going on there.

Yeah, absolutely. So Bandsaw AI, it's a project that started last year. It was a really good opportunity to merge the experience of being an operational leader and just adding AI to it. It was just amazing and eye opening, to be honest with you. And basically what we do is that we help companies understand how their workflow actually flows. We help them identify the highest impact, manual work, and we automate it in a very practical way. Basically, our focus is return of investment, not the AI hype. We have worked with multiple companies from home services all the way up to staffing companies, staffing firms, and some of the common use cases. And the things that we help them automate and get a lot of return of investment in is speed to lead automation, no show prevention and scheduling workflows. We have done some automated invoicing and status updates in multiple CRMs. We have done social media automation and customer support automation. But that is not the best part. Those are the things that we have already ready. After we just understand your business, we can just plug it in. But we also work with custom workflows and automations. So that means that if your process is unique, if what you do is completely different than what everybody else does. So we help you automate that process, we help you map that workflow end to end, identify bottlenecks, and just deploy solutions quickly on your current infrastructure. So I think that's the best part, that we can make this AI implementation only yours, no one else will have it.

You said something that I think resonates a lot. It's 2026, but even the last two years is AI hype. It feels like everything. Every newsletter that I get is about AI. Every investment opportunity is about AI. It feels like everyone is using it, but very few understand it. Tell us, what does this AI hype mean to you? And how are companies maybe like kind of paralyzed by all this?

Yeah, absolutely. When you think about AI hype, that means people trying to adopt AI just because they want to say I have AI in my business without even understanding how AI can actually help them just to get a better return of investment or increasing their revenue, reducing errors even. There are two type of people from what I have learned in this time. First, the early adopters, which are like the visionary people, those ones that are willing to take risks just to get competitive advantage. And the other piece is like early adopters, those are more pragmatic and what they want to do is they want proof, return of investment, repeatable results before committing. So there's this chunk of people that are already making money and reducing costs using AI because they understand where in their workflow, where in their process AI fits. It's not just, you know, just putting one AI tool because everyone else is using it. I think those are the ones that are making money right now. And the early adopters are those again, that they understand that AI needs to be in their business, but they still don't know how to do it. And that's where men saw AI helps. In short, I got to say that more leaders are now willing to try AI and the mindset has shifted from curiosity to urgency. But the truth is just only one. If you don't use AI in your company, then you are at disadvantage, but you got to know where to put it in.

Yeah, can you talk about. I think a lot of people, even myself included, over this last year, I've been trying to understand like AI tools versus automation versus agents. Can you just kind of help us understand like the definitions of that?

Yeah, absolutely. And I want to keep it simple for you when we talk about tools right now in this 2026, there are two different differentiated buckets, like very clear. The first one is the bucket of by coding. These are like software tools where you can create things like websites, internal tools, even mobile applications just by using natural language, plain English. You need to know how to describe what you want and the tool helps you build it out of nowhere. It's going to take like five minutes and you don't have to be a coder or software developer just to do this right. And the goal of this type of tools is just getting a speed and validation of your project and your idea. Personally, the tool that I love is lovable.dev because it has a lot of integrations built in, very easy to use, just a simple prompt. There are other tools in the market, but for me, lovable.dev is the most powerful and the other piece of it is the automation platforms. So yeah, you got it, you got your app by coded, you have the user interface, you really like how it looks, but what happens when you start to use it? So the backend is built in these automation platforms, which are pretty much tools that connect systems, move data and trigger actions for you. My preferred tool is END because it is flexible and strong for building production workflows. There's one mistake that people do and that most companies do. They treat these two as separate experiments. But the truth is that the right approach is treated as one system. You use the bytecoding tool to create the user interface. Very simple, very easy to use. And you connect it to the backend, to the automation tool so that it delivers value at scale. So the rest of the tools that you will see in the market is just a combination of these both. So very simple to understand in that sense.

Marvin, when people say AI agent, like what does that mean? You know, are these like little robots that are doing stuff? Is it really N8N and people are kind of, you know, they're confusing what it is?

Yeah, that's a very common question. Let me just explain the difference. We have workflows, we have workflows with AI and we have AI agents. Workflow is when you have a very well defined process, conditional logic. If then step one and step two, you don't need AI for that. And surprisingly, 70% of the workflows that we build for companies, that's only what they need. Just having a clear structure. Workflows with AI, it's. There's a point in that process where logic is not the only thing that defines the next step, but you need to have like a gatekeeper, something that makes a decision. Like for instance, if you get an email, how do you know what type of email is that? That's when you implement AI in that workflow because the AI can help you understand the content of the email, but that is not an AI agent yet. So that workflow becomes an AI agent when you provide access to a tool to that agent or to that workflow. So, for instance, you got A workflow that identifies the type of email. It classifies the email based on the intent of the person that wrote it. But if you allow AI to, to draft a response, to even send an email to classify in your environment, that's an AI agent. So it's basically an AI workflow with the capabilities of interacting with tools. Now the other piece is how you make it look beautiful. That's where the by croning tool takes place.

Got it, Got it.

Okay.

Okay, good. When I hear about companies that were automating using Zapier, which I think is like the most well known API connector, you know, automation, can you help us understand like in what case is it better to use Zapier? In what cases should somebody look at N8N or something like that instead?

Yeah, Zapier, it's like one of the first as you say, you know, it helps you connect, similar to N8N but because it's been in the market for so longer, it has more integrations. So think about integrations as tools software. For instance, Gmail is an integration, Trello is an integration. ClickUp is an integration. So what Zapier does is just to help you with the connection to all those tools in an easy to read format and easy for someone with little knowledge in technical stuff. N8N on the other side has less connectors, they call it nodes. So if you want to connect a tool that is not built in, then you need to have more technical knowledge to actually connect that tool. For instance, if you want to connect Trello and you need and you don't have all the passwords and API and stuff, so you have to go into the API documentation, look at how to write it down, create an HTTP node and I don't want to go to technical, but in essence it's easier to manage. So if you're learning then Zapier is the way to go. But if you know what you're doing, then NAN is the best tool.

Yeah, yeah, Good, good. Tell us more about the types of customers that Bandsaw is working with and able to work with. Are you looking at small businesses, founder led companies, mid market larger? You know what, like what's the right type of customer for Bandsaw?

Yeah, Banso ideal customer is either a founder led a small business or mid sized business that they want to scale without adding headcount. So basically what we do is to amplify the process that you already have and then just adding AI to make your process faster and to reduce errors. So when thinking about the right customers that we serve Is small businesses, mid businesses in any area. It doesn't matter if you're home services, it doesn't matter if you are a plumber, a cleaning company even, even if you are a high tech company. We can help you as long as you have your processes well defined and you have with your SOPs up to date. Even if you don't have all that in place, no worries because we are can help you map those processes for you. And that's when I get in because I have a like a very nice background of operational experience.

Yeah. Do you think companies right now are maybe trying to evaluate do I need a virtual assistant, say from the Philippines or do I need AI? You know, how do, how might you advise like when you need a person versus when you could use an agent or a workflow.

Based on my personal opinion, you will always, always need somebody who is going to oversee your AI process. But don't get me wrong, you might need less headcount than when you think. So probably you are thinking of hiring two people to manage your social media, just to give you an example. But you can actually automate your social media process your posting, your meta ads if needed and then have AI doing that for you. And then having just like a virtual assistant who can oversee the quality of the output and can start also teaching the AI what's wrong, what is not so good and what needs to be changed. So it's a combination of both AI for amplifying your processes to making them faster and better and just a person that can oversee all that on your behalf.

Yeah. As a founder of a couple different businesses, I can maybe comment on something that you were able to build that I think I got a lot of value out of. In professional services, we believe that consumption drives conversion. People want to look at your content a lot. They want to read it, consume it, digest it and then when they're ready, you know, they reach out. I think the thing that nobody talks about is it's kind of exhausting to come up with content all the time and use the stuff that I've built in the past. So There was a LinkedIn automation tool that you built and I was hoping you could explain that to the audience like what the thesis was and then how you went about building that.

Yeah, it would be a pleasure. So we built this for actually Plot Technologies, which is one of our clients. So what we did is to explore the process. How do you, how do you initially do this process manually? How do you plan for your content? And then what we did is to Use a bytecoding tool, which is lovable.dev to build a user interface. And what's most interesting about this is that we could build like a database of content built from the plug that tech website just to feed our AI agent. So that way the content automation tool would work the following way. Number one, it's going to be triggered by an email that it's running to the founder every day. The founder can access that UI that we built with lovable.dev and from that UI user interface, the founder could, from topics that were already on our website, just to write down a social media post on LinkedIn, because it was specifically for LinkedIn. But not only that, but we also gave this AI tool instructions on how to craft and how to create that post. That post goes to the founder for validation. Again, I think it is really critical that a person can just oversee the whole process and, and if the person, in this case, the founder would agree to it, he would just approve. And that was connected to his LinkedIn profile and the, and the post was going live as soon as he hit approved. So it was just amazing. It was saving a lot of times on ideas and he was saving tons of time per week when it comes to just creating the content.

Yeah, I'll, I'll echo that as somebody who would wake up and then almost feel the stress of like, what am I going to write about trying to get the post exactly right. And then now it's really three clicks and I've got a post that feels like it's in my voice because it's pulling content that I've already created. And it's easy, saves a lot of time. But I also think, you know, there's this kind of too good to be true mentality a lot of people have, like, if it's really easy to do or it saves time, all right, like there must be a drop off in quality. But I haven't seen that. You know, there's embedded videos, clean writing, proper tags. Impressions are still right up there. So I think it's been a success story. It's been more automated, more time saving really, without the quality drop off.

Yeah, and that's, and that's actually the, you know, the end goal, make life easier for business owners and people running, you know, at their companies. We want to give them time back so that they can focus on, know what really drives growth, like taking those strategic decisions in their companies.

I think maybe three, four or five years ago, it was common for, say, virtual assistants armed with a Canva account to create social media posts. And you know, similar to what I went through, there was like a little bit of a stress like what I'm going to write about. I got to make everything proper and clean and pretty. Tell us about how companies are using AI for social media posts.

So the use case is very easy as long as the processes is fully mapped. AI can help you, number one, come up with ideas of content. Number two, AI can help you actually create videos for you and make the connection coming from all the way from the user interface to the automation piece to Facebook, Instagram, LinkedIn. You can actually. Did you know that there's a tool that you can actually post up to nine social media platforms at the same time? And yeah, and that's not the best part. The best part is that each social media platform you can actually write the content differently so you can train the AI to do it. So I have seen a lot of companies just taking pictures of their product, just enhancing the picture and creating new designs just by telling AI what to do. Even creating short videos from that picture from. They call it UGD videos, like people using your product and then just uploading that to AI. And I've seen a lot of companies being successful just getting new followers, getting new likes, getting a lot of traction in their social media with which also creates trust. When you actually go to Facebook, Instagram and you see a company that's been posting consistently, then you say, hey, this company is active and can really drive results. And that's what we do with AI. And that's one thing that Bandsaw can help you build.

Yeah, I think the last part of the process I wanted you to dive into is sales lead generation. You know, all this world and we've seen, you know, HubSpot and all these different tools and Apollo and feels like it's noisy. It's already kind of easy to do that. But maybe there are other ways to do this. More inexpensive but equally as efficient. Can you talk about AI tools and other ways that companies are leveraging that for Legion?

Yeah, and I don't know if I have to go too technical today. This. But you know, this is what I love about AI. How many tools are available now? There is one tool which is called and it's a website which is called Appify. Appify has multiple actors that how they call it that can help you do a lot of things. One of those things is scraping for new leads. So you can actually use those to go into Google Maps, going to Facebook, going to LinkedIn and look for people based on the criteria that you provide. And that's just one tool. Then once you have all that information, you can parse that, or you can move that information from Apify into your backend, which is N8N. Right? From N8N, you can use AI, an AI agent that can actually go deeper into the web. And with the initial information provided, you can actually read research on one specific need. So, for instance, we can go into your LinkedIn profile and look at your experience, your posts, everything that you've done for the last years. And then AI can actually give a summary of you, what you do and the companies. And then we can use that summary to create a personalized communication for you, either email, DM or whatever you need. And then we can connect that to your CRM because we want to keep track of all the leads that we have targeted. So that's the beauty of AI. And that's another way where you can get new leads, personalized messages, and keep them up in your CRM up to date without you making huge efforts. You can pass this list on to your sales agent afterwards, maybe somebody nearshore that you hired, and then that person can start making more targeted calls to them, those people. So there's so many ways that you can plug AI into your business. You just need to know where exactly in your process AI fits better. The key if you want to get a quick return of investment, don't make it complicated. You don't have to go with the hype. Just let's look at your process, let's see how you do things. Let's plug AI and let's get money back into your pocket. I would really love that.

Yeah, I think so. What does it mean right now to be in Latin America using AI? And maybe let me give some context to this question, because, you know, I think sometimes people in the US can feel like they're the only ones on the forefront of technology. Are people in Central America, South America, are they using the same tool? Are they using ChatGPT and Gemini and N8N. Is it, is, is it more rare? Tell us about the adoption of AI in Latin America.

I have a lot of people in my network that are using AI and that are located in Central America. And you can see two different types of people. You can see the people that is targeting the US market, so they try to keep up the pace with the us, but in Latin America, there's the other part of people that are just learning. So when it comes to AI adoption, I hate to admit it, but Latin America is Just one step back, like one step behind than the U.S. but this is not because of lack of talent. There are clear visionaries here, people that understand that AI can really make a difference in their business. The bigger issue here is that most people are still stuck in the curiosity and learning phase. Instead of applying AI into real workflows, maybe they are afraid to move on. Maybe they don't trust 100% of AI. So they are in the process of discovery. I'm pretty sure that once we understand better and start implementing AI into our businesses, that's when the adoption is going to accelerate quickly. I think it's going to take maybe six months to one year before we move on to the next step stage in Latin America. But there's a lot of noise. There's a lot of noise here when it comes to automation and AI, but there's a lot of teaching, teaching courses. That's where we are extending that compared to us, where we are already implementing it.

Yeah, yeah, I think that's right. Tell us more about Managua. You know, what is it like living in Nicaragua, doing business there, working with the us, the business climate, the culture? Tell us more about that.

Yeah, well, living and working from Managua, it's like a contrast between intensity and balance, to be honest with you. So one of the things that I really hate about Managua, it's traffic. Traffic is a real challenge and it can be heavy pretty much at any time of the day. So if I want to go out to the street, I think about it twice, to be honest with you, because drama. Driving just five miles is gonna take me anywhere between 30 minutes to 45 minutes. And I am not joking. So this forces you to plan your day with intention. Because if you do not manage your time, then the day is gonna manage you. And that is so crazy. But once you beat the traffic in the city, right, the lifestyle that you can experience is just another level. In Managua, the beach is like about 45 minutes away after you get out of the city. So it's possible for you to disconnect without doing long road trips. Very easy. There's also a bunch of restaurants. There's new restaurants every day here. And there's a lot of places where you can go to enjoy both personally if you want to take some time with yourself, which is okay, or with your family. Lastly, the cost of living is not that bad. So that's create a lot of flexibility. So it's easy. It's easy for you to build a comfortable life here in the city.

Yeah. Tell us about the food, the music, the culture. You know, for those listening who've never been to Nicaragua.

Well, so food, you will find, like, different traditional dishes. So you got places where you can get American food, like hamburgers, hot dogs, and some stuff like that. We also have places where you can get Chinese food, where you can get food from these countries that are very, very traditional, like Peruvian food in that style. But there's also, like. Like a lot of these places where we sell typical food. We call them fritangas. So fritangas, it's typical food. Gallo, pinto, carnesa. And you can find an open fritanga every day at night, Monday through Sunday. So if you're hungry, then just go out, take a walk, and just go to a fritanga, where you can either eat in place, which makes it better because you are in the street looking at people, or you can take it home and then just be relaxed. But that's the best part of being in Managua, Nicaragua.

Yeah, I think that's great. Well, Marvin, this has been a really educational show. I think making AI easy to understand and not intimidating. You know, not scary. You know, that was a big goal for the show today, and I think you accomplished that. So thank you.

Thank you, Brian. It's been a pleasure to be here with you.

This podcast is sponsored by Plug Technologies, plugg Tech. Great way to connect talent from all over Latin America with the US Companies. Thank you again to Marvin Martinez, CEO and founder of Bandsaw AI. We will see you next time on the Nearshore Cafe podcast next week for.

A new episode featuring another special guest.