What we talked about
Dan Perera explains what JX does in simple terms: they help organizations:from small businesses to large enterprises:solve real operational problems using technology, including workflow automation, dashboards, custom software (ERP/CRM/SaaS), AI integrations, and even blockchain solutions when it fits. Dan emphasizes that every system and process ultimately comes back to people, so JX starts by understanding the human side of the problem before building anything.
Show notes
Dan Perera spent 20 to 25 years working across five continents, Australia, New Zealand, Singapore, Dubai, the UK, and now North America, before concluding that most organizations don’t actually solve their problems, they just keep patching them. His firm JX built its entire methodology around the habit he developed of walking into meetings, staying quiet, and then naming the elephant in the room that no one else would touch.
What we covered
- Dan’s path to starting JX began in corporate boardrooms across multiple countries, where he noticed the same pattern everywhere: committees would spend hours discussing symptoms while the real problem sat unaddressed. As an outside consultant, he had the political freedom to name it, something employees couldn’t do without risking their jobs.
- JX’s discovery process starts by mapping a client’s full weekly workflow horizontally, what happens Monday through Friday, touch point by touch point, before identifying any pain. Only then does Dan’s team ask what the ideal future state looks like, combining current pain removal with a five-year strategic roadmap into a single proposal.
- On the question of trusting AI in automation, Dan described a middleware algorithm JX built that sits between a client’s system and external AI models like OpenAI. The algorithm sends prompts out but blocks data from returning, meaning the client’s data stays within their own infrastructure. JX has tested this approach with 16 different AI models.
- Dan explained why locally trained models outperform general models for business applications: a model trained only on a company’s own data has a much smaller, more specific dataset to reason from, making it increasingly accurate over time compared to a model drawing from the entire global web. He described this as the key to making AI consistent and scalable.
- He pushed back on the idea that AI should be implemented everywhere by default, citing a car dealership where AI call routing frustrated customers who just wanted to speak to their usual service contact. His argument: the right question isn’t “do we need AI?” but “where in this specific workflow does AI actually improve the human experience?”
- His biggest early business mistake was ignoring cash flow management. He went back and completed a short course in entrepreneurship and finance, and now advises every early-stage founder to learn accounting, local regulatory requirements, and data privacy laws before launch, not after.
About Dan
Dan Perera is the founder and CEO of JX, a technology company that helps businesses modernize and scale through AI-driven software, intelligent automation, and human-centered product design. With a career spanning over two decades and multiple continents, he brings both technical depth and a background in public speaking and executive coaching to client engagements.
- LinkedIn: https://www.linkedin.com/in/dan-perera-jxdesign
- Website:
Episode 103 of the PreVetted Podcast.
Full transcript
Federico Ramallo (00:01) Welcome back to the pre-Vered podcast where we spotlight extraordinary people and remarkable talent reshaping our world. Today we are joined by Dan Perera, he’s founder and CEO of JX where he helps businesses modernize and scale with AI driven software, intelligent automation and product design. Dan’s work spans everything from custom
apps and dashboards to AI workflow automation and even blockchain token-based systems with a clear mission make powerful technology simple scalable and profitable in this conversation we’re going to talk about how he picks the right problems to solve what AI transformation looks like in the real world and the practical lessons he’s learned building systems that drive real outcome Dan welcome to the show
Dan (00:54) Federico, you’ve been very kind of the great intro about me. I should say it’s a privilege to be here Federico and talk about what we do and how we shape our world. So I’m excited for this session and yeah, let’s start.
Federico Ramallo (01:09) Awesome, awesome. I’m super excited to have you here today. We’re honored to be able to share your story with our audience.
Dan (01:17) Thank you.
Federico Ramallo (01:18) So let’s start by talking about what does JX do in simple terms.
Dan (01:26) I think you kind of had a very good lead into it. That’s exactly what we do. So we are an automation company and we…
It can be a very small problem or it can be a very large problem for any enterprise. can be a small, or let’s say that you are a very small entity, small business, right? And you can be an enterprise that has about thousand employees in it. What we always see is problems are problems and it’s related to people end of the day. So no matter what type of a system or what type
of a workflow or processes that you have, it all comes back to people. So the way we approach is basically understanding that. So our company is really focusing and honing on
the problem statement, right? So what we do is we solving problems with technology. That’s who we are. So it can be, like you said, it can be a simple automation, improving workflow with a simple solution, an integration, for example. Or it can be, we have multiple softwares running vertically and…
I mean copy paste so and so let’s build a dashboard bring everything together then you solve that problem right
And then we get asked, how can we adopt AI into that? So does that mean that we are building a brand new or from the scratch and model that helps them, or we actually bring in an existing AI model to run through that integration as well? And that’s another solution we can provide. And if not, you know what? We’ve had enough of all the systems we had. We wanted to build like a brand new
enterprise resourcing program like ERP or CRM or any other software. So we then customly build that automation software as well. So that can be within the web too, which is software as a service, a SaaS product. It can be a mobile application. It can be a dashboard integration, and it can be a complete AI model.
a complete AI automation or like you mentioned that we’ve been working on blockchain technologies as well, especially in the financial sector. So we kind of looking to that as well. So end to end, we pretty much cover from a small business to large enterprise on their problem solving mechanisms.
I just wanted to add this into that mix, which is we believed in human-centered design. I’m a big believer in it. So we put humans in the middle of a problem statement as the solution.
Federico Ramallo (03:58) Amazing, When did you realize you wanted to build a company around software and automation?
Dan (04:06) I’ve been in the industry for the last 20 to 25 years now and I’ve been in the corporate sector, I’ve been in a small business sector, I’ve been in a startup, I have worked in multiple countries like of course Australia, New Zealand and Singapore, then Dubai, the Middle East, UK and also now North America.
Before I started my company, I left my corporate world and I started consulting. My father was a businessman, I think, it runs in the blood. I realized that, like, going from one company to another, working, and even when I left and started consulting,
Like nobody wants to solve the problem. Everybody wants to patch it. Everybody wants to put another patch and another patch and another patch. Like you sit down in boardrooms and talk about it. Like you’re just wasting our time talking. Nothing gets done. And I’ve always wanted to be a problem solver. I always wanted to kind of…
point the finger at the right direction and ask the difficult questions. I’ve always been that. And every time I walk into a room for a meeting.
I don’t, I don’t say much at the start and I wait until the end. They’re like, damn, what do you think? I’m like, this is the problem. But you guys are going around in circles, but not talking to the elephant in the room. Let’s talk to the elephant in the room. That’s my job. Like you’re just talking something else, but it’s difficult. It’s very hard, but that’s where, that’s where we need to address. And it gave me like, okay, I’m consulting now. And people started listening to me. Of course there are nuance of psychology.
Federico Ramallo (05:26) you
Dan (05:42) behind it because you’re an external party. You know that like our kids never listens to us, right? But somebody else says something, they listen to them. So it’s like exactly. So when somebody comes outside, people start to listen. So I decided like, this is a very good way for me to make that impact into companies. that also gave me that drive. Like I said, I’d probably been somebody who have that…
Federico Ramallo (05:51) Yes.
Dan (06:07) ambition to become entrepreneurs there. So there’s a mix of a lot of things.
Federico Ramallo (06:12) Amazing. Yeah, I think that that’s… I share the passion of solving problems as well and just not patching it. And there is some beauty on finding the optimal solution and optimizing, you know, because there’s always trade-offs on a solution, right? But over-engineering a solution that, you know, is just efficient and it works and it’s solid and it’s, you know…
and you see it running and you know that for me that’s amazing right and being able to answer all the questions saying you know why didn’t it it’s already solved you know but have you considered this use case it works perfectly for that use case you know
Dan (06:44) I agree.
Hahaha.
Yeah, you’re right. Absolutely spot on on that.
Federico Ramallo (07:01) Yeah, yeah, You run the continuous integration server, everything looks green, you know? It’s such a beautiful thing to happen, right?
Dan (07:12) Yes.
Federico Ramallo (07:13) Yeah, yeah, yeah. We deliver a feature, you know, a new feature. We launch it on time. You know, we don’t have any bucks, any big bucks, you know, coming back. Right. It means that we did a great job that we build something that it’s, know, yeah, yeah. I mean, for me that that that those things are, you know, beauty per se. Right. Yeah. And and I agree with you that people listen to outsiders more than within the organization.
Dan (07:33) yeah, you’re right, absolutely right.
Federico Ramallo (07:42) There is some, I think that there are a few reasons for that to happen. One is that once you’re within an organization, you start building these relationships and these status quo. So then people are afraid to break the status quo, right? Or challenge it, right? When you come as an outsider, you have that freedom of just, you know, going…
challenge everything because you’re new, you’re asking questions, right? And you’re in that process challenging the status quo. And you can ask, why are you doing this? because that is how it’s been done for whatever time, right? And you can challenge that without any political expense on your behalf, right? But if you’re within an organization and you’ve been there for many years, then…
you know, people kind of tend to go to don’t rock the boat, right?
Dan (08:33) You’re right. You’re right. That’s see, I’ve experienced that when you’re an employee and then this like I said, it’s very psychological, right? When you’re an employee and it goes back to the basics of your existence, like, okay, this is my job and I don’t want to rock the board. had to protect my job. like, even if you want to say something, you probably think twice about saying that there’s so much nuances around it. Like, you know, even you want to hire
like a position, you still kind of play safe sometimes, but when somebody from outside is coming, you don’t have that responsibility. So you’re basically saying the actual agency we hired to do that, they told us to do this. So now you’re pretty much…
putting the argument, evidence-based arguments. So evidence has come through a third party. So it works. Like I said, what I’ve experienced across my career, I mean, including when after I started my business, is that how can we communicate and how can we bring the conversation to support those members on the board or those executive or decision makers that…
they will use our advice or our information that brings to the table as an evidence. So it becomes easier sometimes to make those decisions as well.
Federico Ramallo (09:54) Right, right. I wanted to ask you something about automation. It’s so how can we build a system with AI automation that we can trust given that AI is non-deterministic?
Dan (10:10) I’ll give you a little example. So I’m going to talk about something that we are working on. It’s almost at the implementation level. It’s one of the products that we are building right now.
Federico Ramallo (10:13) Yes.
Dan (10:20) If you look at like this is year and a half ago 2024 August I think I was I was talking to I’m a public speaker so I do keynote speeches as well so sitting in downtown LA and I I was asked to give us a speech about AI because everyone was like my god that’s that’s the only thing that it was just exposed to. However I was kind of before I spoke to the audience I did a quick
research on but what’s out there. This is a year and half ago. There were about 125 AI models came out within that week. That’s a year and a half ago.
Now, if you go and look today, I’m sure there’s about a thousand odd models are coming out this week or last week. Simply because AI models are not hard to build anymore because there’s so much, like we can talk about Claude, for example, right? So Claude is helping you to build more and more of it. So it’s becoming, the bell curve is coming to a plateauing, like, you know, that just like,
late I would say in 2008, 2007 and 8 we went through the SaaS boom like there were big companies doing SaaS stuff everyone’s talking about it okay how can we build another SaaS another product and another product and another product so there was a boom around it then it’s it became expensive it just became viscosity then it becomes normalized so we are going through that phase of AI as well right there was a huge boom of it still booming but now from automation like companies like us we see that it’s flattering it out now it’s just
balancing the industry around it. Now the trust, I want to get back to that question. Now because there’s so much out there, right, then there’s no regulation. It’s a wild wild west right now with AI. Anybody can do anything and put it out of the market and use it, there’s no regulation in place yet. However,
Another reason is that what we are building right now, I’m going back to where I started, that we manage to find, we manage, rather, we are confident, we manage, we’re confident, we manage to find a, we manage to find a method as well as a middle valve. We run another algorithm in between. Now, let’s say that somebody wants to build an AI automation software. Does that mean that, does everything run only on AI? Absolutely not.
It’s not. So AI only doing certain parts to the application. You still need to have data entry, right? Because without that entry, AI can’t do anything. Without data, AI can’t do anything. So the human-company interaction has to happen. So when that happens, these automations that’s pulling in the AI automations into it, right?
always the model, let’s take OpenAI, which is Chatch-IPT. Let’s say we are bringing the API out of the OpenAI Chatch-IPT into it, right? Generally, they have the settings. You can actually tell them not to share our data with OpenAI. Now, does that mean that we’re going to use it? I don’t want to say it. I’m just going say, don’t know. We don’t know, right?
Federico Ramallo (13:36) You
Dan (13:39) We build an algorithm. We build an algorithm as a middleman now.
We are bringing the API from OpenAI to run the requests, run the prompts within these dashboard or the platform that you’re in. And it only runs within you. It doesn’t send back any data because we are stopping that API to go back. It’s not going to happen both ways. It’s only happened one way. However, every time there’s a prompt being hit into that particular application that we’re building, right, let’s take it as a dashboard or whatever.
It’s going to call. It’s there’s going to be a call from the system that we are building to the open AI to ask you, OK, system is there’s a problem coming in asking this information. Now I want you to run that logic or the intelligence. But right, you know, then it’s using the data within the database of the customer only and then give that answer or the output just onto their side. It’s cut. This is very complicated, but I’m just trying to simplify.
as much as possible.
So there is a way, there’s a way to actually create AI automations without creating your own model by using existing models. But there is now, there are ways to stop data being kind of moving back and forth. You can actually keep your data within your database, within your system, but run the model within your application without getting them access to that. There are ways to do that. So we are building that and we have kind
of almost we already built it we tested out its works it’s not only with OpenAI we have about 16 other models that can be done with it.
Federico Ramallo (15:20) very interesting. Yes, the way that I understand it and I apologize if I, you know, kind of badger the concept but…
Dan (15:26) I was
gonna ask, does it make sense?
Federico Ramallo (15:32) Yes,
I think it makes sense. I’m trying to rephrase it to see if I understand it properly. ⁓ So each model has a data set, lots and lots of data that put it together. That’s kind of how the model is built. And now, as you mentioned, there are variants of other open source models and data sets and combinations that you can make different.
Dan (15:38) Yeah.
Federico Ramallo (15:59) models for different things, Whether it’s, know, but just by changing the data set and then you can also change how the model behaves and how it interprets the data, right? So, and I’ve seen, yeah, I’ve seen so many variants of that, right? And then from with that, and yeah, probably this is a little bit beyond of what you described, but then you have the same model with different size of data sets.
so it can use more or less resources when it’s running, right? Dimensions, I don’t remember the name, but yeah. And then you have what I was reading about context where you can say to a model, well, this is the context that you have in order to make, to build a response, right? And that context is what you were describing as…
the information, the local information of the company or the project or the task, whatever it is, right?
Dan (16:55) Yeah, so if I put it into a simple analogy, let’s talk about a business agreement, for example, right? My business, your business, and somebody else’s business, right? Now, I am the owner of my business. That’s where need my operations has to happen. And I’m going to tell you, right?
and I’m gonna write a contract with you that we are gonna do business together, right? But I’m gonna tell you this business called ABCOXYZ business. I don’t want you to tell them everything that we are doing.
Just tell them that we are a business of selling caps or selling water or something. That’s all they need to know. They don’t need to know how we are going to make the bottle, what materials we are going to have, how the water has been recycled. That’s only you and me. That doesn’t need to know all that stuff. So the contract we are going to sign with you and the contract that you’re going to sign with them. There’s so much. So this much of information we both of us know. This is only they need to know.
So, in a nutshell, then the API basically, you are the middleware algorithm. We are the customer that we are running the software and the XYZ business is open AI. They don’t need to know everything that we do.
Federico Ramallo (18:18) Right, right, you can define the scope and you can define how much outside of the scope of knowledge do you want the model to walk outside. Are those hard limits or soft limits, right? ⁓
Dan (18:32) Absolutely. And also
there are this like also there’s a few other platforms like they are like N8n is something that is really, really good at model or like a model customizations. So when we do customize models, so we get existing model, if we customize it, we can pull that algorithm as an individual model and then without any API, just implement into your software. That’s fine.
That way the data is not going anywhere. You’re running the model within you, train your model within your system. So it’s everything sitting in your side.
Federico Ramallo (19:05) Right, right, and nothing leaves your infrastructure.
Dan (19:09) Absolutely not.
Federico Ramallo (19:11) Have you seen better results with a local model than… Yes. Amazing. Yeah.
Dan (19:15) Yes, simple.
Because I’ll tell you why. Now, I’m sure these words are now around in the industry and market as well. So there is a stage called hallucination state of any AI model, especially on the large language models. Now, hallucination is a simple word that
It’s the exact same meaning when we hallucinate as well as AI hallucinate. There’s no difference. So hallucination is simply like, let’s say you’re going to ask me a very complex question, Federico. I’m thinking, I’m thinking, I’m thinking to answer.
I’m answering this in a very simple way so everybody who listens to the understand that. So I’m thinking, what is thinking means? It’s not hallucination. Sometimes you go into the stages like, is this right? Is this not? I’m like all over the place, right? So I’m just trying to figure it out, right? You’re in a stage of hallucination, a stage of moving your mind all over the place, right? So…
This is exactly what AI is doing from an LLM perspective, trying to figure it out. It goes and then thinking, it says, I’m thinking, I’m thinking. So thinking means it’s actually in that stage of hallucination. can, so it can narrow it down and bring it back quickly, or it won’t be able to narrow it down and come back quickly because it’s looking for vast datasets in the world. If we take open AI and then give you the best solution. Now in this state, hallucination can got really vivid.
and then won’t be able to provide you the right solution. probably, everybody probably have experienced that at the start of any of this OpenAI to any other, it couldn’t give you exactly what you’re looking, but it will give you something, right? So that’s what happens in any large language model until you train them to become more accurate and accurate because more data sets have access to it. Now, localization, right? That means your model only runs within you. That means now your data set
become very small compared to the worldwide web of like where you have access to say that OpenAI have access to pretty much every single server in the world for example right in localization you only have access to one server that means the more you run that local model
it’s going to get much more smarter about your data set than any superior model like OpenAI. Why? Because you’re not going to data sets in Europe or Asia or anywhere else. You’re only sitting in your company. Now the model is becoming smarter and smarter and because it’s understanding the data sets very clearly.
Federico Ramallo (22:03) Right, right. mean, it’s I don’t know if this analogy applies, but I hope it does. But is this idea of there is this concept of ignorance is bliss where, you know, if you if you have so much knowledge, then you have so many ideas and thinking about so many things. But if you have, you know, if you you think the sixth day, you know, movie with just Schwarzenegger, you know, where they make a clone and they
they transfer the knowledge that that person had at the snapshot of that, but doesn’t know anything after that. So they can make a great clone of the character. Or there’s this other movie where they manufacture soldiers, and they tweak the memory where they only know about, they become the perfect killing machine because all they know is how to become the great soldier.
Again, in the movies, you know, it’s sci-fi, right? But what I’m trying to go with this idea of if you have a very specific subset of knowledge, then your answers are going to be specific to that, right? Whereas AI is trying to be this generic know-it-all thing where it is more likely to give you the unexpected answer. Maybe it’s not wrong, but…
in the context of a company, I just want the information of the company I don’t want you to think about life and movies and whatnot, right?
Dan (23:26) Yeah, pretty much you’re right there. it just, I would say the word would be localization and like, know, narrowing down to single source of datasets, your model becomes much more accurate. It’ll become much far superior than any other, any other models out there.
Federico Ramallo (23:41) Right, right. I was able to run local models in my laptop, got in a plane and ask recipes for chicken, right? How can I cook a chicken, right? And it will give me very simple answers, right? Because it was one of these low memory models, right? So didn’t have much information, but it was good enough, right? But then with the larger models, I was able to, you know…
I was able to get a bigger computer with 64 gigabytes of RAM and whatnot and I was able to build around larger models and ask the question and then it would ask me for nuances, know, how do you want the chicken to be cooked, right? Which is a more, you know, complex reasoning, right?
Dan (24:29) Yeah, absolutely. You’re right.
Federico Ramallo (24:34) So I think that the balance is to find how much specific you can build a model and how you can provide the specific data so you can provide much more specific answers, right?
Dan (24:47) Correct, absolutely, yes.
Federico Ramallo (24:50) Very interesting. So you’ve said that your goal is to make tech simple, scalable, and profitable. What does that, each of those means to you?
Dan (24:58) Absolutely.
Yes, so I’ll give a little context away from every kind of left off of that, know, know model being accurate model being actually consistent and model being scalable. So when you again going back to localization of you know, we have a model that
in your system for your business only and what happens is like the exact words that you mentioned how can we you know make it consistent and scalable and profitable. Now accuracy comes from consistency when the model is getting trained constantly and you know amount of tasks that you ask the model to do it’ll like learn so next time somebody asks that it already done the job it’s going to get better and better that consistency is going to improve and of
that’s helping a business of scale because now the accuracy of the information that provides will give them to make the correct and the feasible and viable.
and also desirable kind of decisions for your business and for your customers, right? And then that’s what makes them profitable because your scalability has because you’re profiting already but your scalability is actually making more profitable because now that consistency is the one that runs through the other two verticals.
Federico Ramallo (26:16) interesting, interesting. Yeah, we used to think of these general models and then how can we restrict it to give us a more predictable answer or more accurate answer. But with
Dan (26:27) Yeah.
Federico Ramallo (26:31) working on building your own model then you can optimize it and you can reduce the cost you can improve the speed to response the accuracy of the response and the impact that that can have on the business can be much bigger and with less resources make it more efficient
Dan (26:49) Correct? Yes, that’s right Federico.
Federico Ramallo (26:51) Right, amazing.
and
So, what are your common workflows that you see waste time and should be automated first?
Dan (27:01) That’s, you know what, that’s a very good question. So everyone who asked that question, I’m not, let me rephrase this.
That question is being answered with a cookie cutter answer mechanism. like everybody answers the same way, like because a cookie cutter question, if that makes sense. Now that’s a very broad question, to be honest. The reason I’m saying that every single business have unique problems as well. So what we do is when we walk into any business, right? What we initially ask is we are not going to touch anything without having a discovery session.
Our priority and the initial goal, we’ll have the initial consulting. We don’t charge for initial consulting, but then we start with discovery session. What we do in discovery session is we are actually mapping the entire workflow of their business. So it’s like a horizontal journey. It’ll give us a journey map.
What’s your workflow looks like? Tell me from Monday to Friday, eight to five, what do do from Monday to Friday? What does that look like? What’s your workflow looks like? So we’re gonna map it out first. It’s a horizontal journey again. After we do that, we actually understand each of the touch points of that workflow. We vertically understand what are the steps you take every touch point, every action you take. There are multiple actions on that vertical.
When we do that, we get a very clear picture of your business model. What are you really doing in your business? Now when we really understand that visually, then we’re going to ask them, OK, how do you do that with that and that and that?
Or we just copy and paste or we just have to scan and upload and do this, do that. Then you’re like, what is your biggest, what are you? What is your biggest pain point in that action? Or what are your biggest pain points across? So now we are actually asking them, what is your problem? What are your pain points? What is really painful for you to do? Right?
Then that helps us to understand one of the biggest problems they have. Now, by doing the discovery session, we are prioritizing what are the low-hanging fruits that we can immediately solve and what are these bigger long-term solutions, long-term problem-solving, you know.
approaches that we need to take. Now when we take that information, we go back and then we create a business model canvas. We create a customer journey map or a business journey map to understand the workflow and then we also during that discovery session we also talk about now we understand your problems and your methods and your process and your model. Let’s say that we really
eliminate these pain points and problems for you in your workflow. What is your ideal state look like? What’s your future state look like? This is your current state. This is your future state. They’re like, oh, we wanted to, we want to do this. We want to become that. We want to do this. We wanted to have another product line.
all comes out. Okay, now we know where you want to be or where you need to be. then what we do is your current state, eliminating pain points, creating solutions for these pain points and then targeting where they need to be in the next five years by strategizing, laying down the roadmap, excuse me, and customizing the solution to reach that goal. So we actually prepare that documentation like I would say,
it’s a proposal, partially it’s the discovery report and of course whether they wanted to go with us after that to really create that automation platform or software or a solution or whether they want to go with someone else doesn’t matter but we always do that right we always do that that’s how we approach now that is helping us to decide do we really need AI now I always say yes you need AI right
now because if you don’t what’s going to happen is your competitor who is using AI is cutting down the processing time from 70 % to 30%. That’s a 40 % advantage that company has over you. So of course AI will help them and of course AI has to come but when we start that’s our first and foremost approach to do.
Hope that makes sense and helps you. Yeah.
Federico Ramallo (31:19) I think that’s amazing. mean, you said something that resonates a lot with me when you are asking, do we need AI? Because now with all the hype that AI has, people go by default to AI first, whatever. And sometimes, you know, I think it’s important to have that introspection to ask ourselves, is this the right solution for the problem?
right? Which at the end of the day could be most of the time, yes, but asking ourselves that question allow us to reason, is this really necessary?
Dan (31:53) Exactly. I mean, it’s very true. Now, do we really need AI or not is a good question. I mean, I have multiple examples I can bring in that I deal with my my in my personal life, I deal with a lot of things. recently I spoke to somebody that who was in a car dealership who
takes care of my car. And then apparently they are trying, he was saying that, don’t call the actual dealership number. It goes to AI automation thing. This is my personal number and call me for anything because it might not come to directly to me.
That problem is not a massive problem. It’s a very tiny problem, but it is a huge impact on me as a customer and the customer service. Now, if I’m going to call the dealership or the car service, then I’m going to hold and this AI kind of agent is talking to me when I know it’s an AI agent. like, can I, I just want to talk to John. Like, can I, can you put me to John? I don’t want, I’ve already done my things, you know? So there are still gaps in AI adaptability.
AI adaptation. So because we are in early stages of these adaptations, we really need to find out where the humans thrive with human interactions and what the customer service is all about, where AI actually can improve the customer experience through the human interaction, not that AI to have a direct interaction with the humans, where humans probably like, why am I talking to a machine? I mean, remember in 2010,
Federico Ramallo (33:25) you
Dan (33:27) I want a chatbot, I want a chatbot, I want a chatbot, I want a chatbot. Nobody then, everybody end up hating chatbots, right? So, because they know that it’s to reduce the amount of human, like I would say, I’ll give you an example. Let’s say that you’re running a business and then you’re service driven and people call you to ask questions all the time. Sometimes they are very silly questions, which you can actually check it on your website. So people are being lazy, pretty much. Now, yes, to minimize that,
incoming calls, having an agent is really screening it out. Totally understand. Now, but if it’s something more complex and if it’s more needs a human solution, human interaction, human to human.
We need to really address it as a human. It requires human to human interaction. So there is still a gap there. getting back to it, just, do we need AI? Yes, we do. But where do we need AI in your business flow? We should not implement AI into any, oh, you know what AI is doing everything. don’t do anything else. Yeah. Yeah. Probably in a nutshell, there are certain businesses definitely can leverage AI into it. Absolutely. No problem at all.
But it doesn’t mean that every business needs AI from end to end. That’s why I’m saying discovery, that discovery session helps us to advise our clients, say, please don’t do that. But if that’s what you want to do with, go ahead with. But you’re going to have more friction with your clients coming into you saying that you.
you know, you have to deal with this separately because our demography that when we’re dealing with our customers, can be an 18 year old, can be a 78 year old. You don’t know that, right?
and but they’re both paying dollars they’re both paying you right so does that mean that you don’t care about the 78 year old when they’re paying the same thing no so you have to figure it out what’s the best model what’s where does it really needs to scale but where does it doesn’t need as much then you need to make those right right decisions for your business so that customers at end of the day if you don’t have customers you don’t have a business so that’s the essence of it
That’s why I said human-centered design is a key for even AI take over lot of things. We still need to think about how can we put our humans center of our solutions. So then we know that ethically we are actually moving forward.
Federico Ramallo (35:54) Right, right. And the companies has been using automation even before AI, if you remember the phone voices, you know, 20 years ago, right? Just to, you know, they would use automation to kind of push away, you know, customer support, you know, users, right? To scare them away or to, you know…
Dan (36:04) Yeah. ⁓
Federico Ramallo (36:16) make it more difficult for you to talk to a human, right? Which you can use technology and particularly AI for good or evil, right? I mean, yeah. So I think that it’s important what you’re saying, human center actually solving, actually adding value to the users, adding value to the business.
Dan (36:19) Exactly.
Yeah, okay.
Federico Ramallo (36:38) through the users, if you give a good experience to the customers of a business, then they’re going to want to use that, are going to provide more business to that business, right? And that turns into profit, right? So I think that there is an important value for that. And as you say, during your discoveries, can decide, you can…
provide recommendations on where companies can use AI or any automation to improve the experience of users and increase profit.
Dan (37:11) Absolutely. mean, I don’t want to contradict myself here when I mentioned that there is like AI, AI will definitely, AI will become fast period to where they, where AI is right now to have a conversation like this very soon. It will happen. However,
I don’t want to predict or I don’t want to talk about like we are age of Armageddon with AI. That’s not what I want to say. But AI will definitely become far superior than where it is right now to have a very sensible and emotional, I don’t know how far that will go with emotions and feelings, but very sensible conversation. It’ll come to that point because at the moment, AI is still learning on those things.
come to it but even if it comes to that I am a firm believer we should not eliminate the human factor out of the equation at all. Does that mean that we are not thinking about humans anymore? You know what I mean? So the human factor has to be there.
Federico Ramallo (38:17) Right, right. It is one of the things that I appreciate of having these conversations because we can learn in a much more intimate environment. We can talk about the nuances of the problems that we’ve how we’re overflowed with content and notifications and whatnot. So our attention, you know.
becomes reduced. So being able to have these long form conversations where we can discuss these topics in much more detail becomes, I think it’s going to be more and more important with how AI is impacting the world.
Dan (38:56) absolutely, I totally agree. I totally agree.
Federico Ramallo (38:59) Amazing. So what is one mistake you made early in your career that taught you a big lesson?
Dan (39:07) career as businessman or entrepreneur or as a tech individual like a designer or engineer?
Federico Ramallo (39:07) You
That’s interesting because you have so many profiles, right?
Dan (39:21) Okay, let me break it down for you then. So as a businessman, as an entrepreneur, the biggest mistake or I would say…
Federico Ramallo (39:22) No.
Dan (39:29) Yeah, because mistake I learned. The biggest challenge and a mistake I made and I learned was understanding finances, understanding how to run a business with cashflow, finances, accounting. I hate accounting. hate anything to do with the finance numbers and stuff. like, I don’t want it. Like I used to kind of always push that away.
But then I, and within the first two years of my company that I’ve learned a huge lesson, we have clients, money comes in, money goes out. Where’s the money? Like it’s so I was having cashflow problems. think all these problems that, but the company’s running. I’m like, what’s something’s wrong.
and like so I went back I rather went back I did a short course learning entrepreneurship and finance so any entrepreneur out there I would say most any entrepreneur out there I would say first and foremost before you start your company or while you’re starting your company learn about
Learn about finances, learn about your cash flow, learn about how to manage money, learn about your regulations, then legalities, per state, per country, per region, like European Union, for example, then Americas, then East Asia, wherever you are. Learn about your state, your country, and your region, and what are the regulations and stuff, what are the data privacy laws are there. So all this comes to everything.
matters to you. So learn about them. Luckily, because I was doing consulting in multiple places that I learned all that before I started the company, and I knew that I need to learn more about it. But the biggest mistake or biggest challenge that I had was about understanding financials on a company. That’s very, very, very important for a businessman. That’s that’s as a businessman. But from a career perspective, I think I my biggest, biggest
instead of rather mistake, again a challenge early in my career, I didn’t realize my potential. I was playing very small and I thought that I was only meant to do certain things. And then you had your managers and superiors you report to and they always tend to box you into something that because they are intimidated by your capabilities. Not just me, I’m talking about every single person in the world that
your capabilities are much bigger than what you think you can do. Everyone’s capable of anything if they put their mind into it. When you know you are capable, but if you’re trying to box you in by yourself as well, by others, don’t let that happen to you. Always think you’re meant to do bigger things. And I think early in the stage, I wasn’t thinking like that. I only realized it later in the stage of my career before I started my company. And I’m like…
I don’t have no regrets, but I’m coaching, I do a lot of coaching on a platform globally. So a lot of juniors, junior tech, tech kind of individuals like engineers, designers, they book me to pretty much coach them. And what I tell them is that I’m not here to teach you about design. I’m not here to teach you about programming or.
business analysis or whatever. I’m here to change your mindset. I’m a coach. I’m not a teacher who’s teaching you a subject. So I think that anybody who’s either you’re in the early stage of your career, doesn’t matter which industry, especially if you’re in tech and if you’re an entrepreneur, even a business owner, like in your growth stage, like always remember these things.
Learn about finances, accounting, regularities, legalities, all of that at early stage of your startup or on your business. And as an early career, if you’re on an early career stage, think bigger, work hard for for that thing bigger. Even though your mind probably says that, you are meant for do bigger things because our parents always say that you’re meant to do bigger things, which is true. But don’t limit yourself, which I did.
I, I, even though I have no regrets, but I wish I’ve learned or I did certain things early in the stage of my career than later. I’ll give you one example. When I start consulting, my client actually woke me up saying that when I speak English,
Sometimes I speak so fast that they can’t understand certain words and certain things I say. So I used to speak much more faster than the way I’m speaking right now. And the certain words that I pronounce, I pronounce that way that they don’t understand.
How am going to change this? Like, you know, it’s like, this is how I speak. So I actually then booked a, like a speech, I went to speech therapy and public speaking coaching and I paid a lot of money to,
to pronounce certain words that people will understand me better. And the tone of voice when you say certain sentences, what that, so if it’s dangerous sentence, then how you kind of express it. If it’s a happy, if it’s a sad one, you know what I mean? So I’ve learned a little bit about, and those things only happen a little later in my career up to now. So I could have learned all that early on that helps me to even progress my career.
Federico Ramallo (44:40) Right.
Dan (44:51) or progress my success early on. things like that, it’s a minor, some of them are major, but I just gave an example of it. Yeah, it changed the way people see you and listen to you. So it has a huge impact.
Federico Ramallo (45:05) Yes, I completely agree. mean, English is my second language.
hated learning English, you know, at one point I hated it, you know. But then I started working with American companies, I started, it was hired by Microsoft when I was 16 years old. So, you know, in those early years, I eventually had to learn English, right? It was one of the things that make a difference, right? And then after that,
when I thought I was like, okay, I know English and blah, blah, blah. I realized what you said, you the nuances of how you speak and how you pronounce certain words. Be able to articulate becomes important, right? And I’m kind of overemphasizing here with my, with how I express it, but, but it is important because people take tend to,
Dan (45:47) absolutely.
Federico Ramallo (45:56) judge the book by the cover, right? And then if they have, you can have a great idea, but if they cannot understand what you’re saying, then they’re going to dismiss you, right? So being able to articulate those ideas become important. And as you said, being able to have the courage to get out of your comfort zone, have the…
Dan (45:58) Hmm.
Federico Ramallo (46:18) security that you know what you’re talking about and be prepared, right? Because all of those things become important because we are the result of every decision that we made throughout our lives, right? And the mistakes that we made, right? So being able to have that security and project it so your ideas can be listened and…
you can articulate them correctly so you can influence others that your idea is a good idea, then that will allow you to grow much, much faster in your career, right? So I completely agree. And then I have been through my years appreciating more and more the different roles in companies, different skills. I used to…
hate financials as well but out of necessity I had to learn it and now I appreciate it more my wife and co-founder she’s running the finance of the company and she does a great job but I also had to learn myself a lot of the concepts to make sure that we are on track right and being able to understand to build a profitable business has a significant impact on
not only to be able to serve the clients, but also to be able to serve the people that you hire, right? Because they decided to join you and your company and they put their life into your hands, right? So then the finance becomes much more significant because you are affecting their lives, right? And, you know, not last or least, right? It’s yourself as well, right? Because if…
you are stressed out because of the finance then you’re going to not be able to serve your clients properly
Dan (48:02) Absolutely, I totally agree.
Federico Ramallo (48:05) Amazing. Dan,
I think that was amazing. We had a great conversation. We’re running out of time. I learned a lot. Any final remarks before we wrap it up?
Dan (48:13) I totally appreciate giving the time of Hedricko for this session. I think we had a really great insights out of these conversations. And I think every time you speak about what you do and also how it’s impacting the world and impacting your surroundings, it’s really, really important and it’s really rewarding. So first of all, I wanted to kind of thank you and appreciate you setting this up.
But from a final thoughts perspective, AI, if you’re really focusing on AI, AI is to stay, it’s not going to go anywhere and it’s only going to grow, it’s only going to dominate in certain parts of our industry as well as the rest of like any other industry as well. So society is going to get impacted by that. And people are worried about is that going, am I going to lose my job?
going to, you know.
Where will I be and whether my job is going to get replaced by AI? Most of the jobs will be getting replaced by AI. It doesn’t mean that it’s going to get replaced completely. It means that it’s helping to speed up the process of what we do. And we look, we need to look at it that way right now. So as an employee or as somebody who is actually trying to grow yourself within that space, would say, learn how to use your tools within AI, just like in early 2000s.
or late 90s, everybody had to learn Microsoft. It’s a very similar impact right now. Learn tools that actually does what your job does. And then you become better at it, then your value goes up. That’s my advice to people who are within the space of employment. For businesses out there who are actually implementing AI into your businesses, my biggest advice, I would say the only advice is
do not adapt any AI simply because it’s AI. Make sure that you do your due diligence on finding out where the biggest problems are in your business and try to adapt AI to do those things. And also, if there are processes and also have workflows that takes a long, long time to do that job,
That’s why you need AI to do your work so that you can actually focus on growing your business up like, know, like sales marketing and then, you know, go and speak to more clients while AI is doing some of these grant work with your employees as well. So always remember, adopting AI the correct way is the best thing to do. Don’t just adopt simply because I can use an agent, right? know, and you’re gonna hate it and your customer’s gonna hate it and you’re gonna leave it.
So that is my biggest, biggest, biggest advice to you because there’s plenty out there to do it. But make the right decision before you getting into AI. Humanize it, right? Humanize it. Put yourself in the middle of the solution. Put your customers in the middle of the solution. And then if you have employees, put them also in the middle of the solution and then bring AI into the middle of that. That’s my advice to you. Adaptation.
Federico Ramallo (51:14) Amazing Dan, thank you very much for joining us today.
Dan (51:17) thank you very much, Federico, and I appreciate your time and invitation. Thank you.