Stop Building Apps. Start Building Agents.

0:00Everything that is not agentic these days is literally legacy work. It is outdated. It's not the way you solve problems today anymore. It's a shift in mindset on how to approach the problems you have in hand.

0:13Matt Paige:Welcome to the Talking AI Podcast, where we talk AI with both experts in the field and early adopters. I'm your host, Matt Paige, and we're here to demystify AI for you so you can get some value from it. Let's talk some AI. Hi. Low code and no code was supposed to be the revolution. Drag and drop, no developers needed. But what happens when you don't even need to drag and drop anymore, when you can just describe what you want and an app appears? And what's happening to every company in that space when you've got Anthropic, OpenAI, Lovable, and a dozen other vibe coding startups all coming for that same market?

0:48Matt Paige:And today's guest sits right in the middle of that collision. He's the CIO of one of the biggest low-code platforms on the planet, OutSystems. And instead of playing defense, they're actually building something really cool. Welcome to the show, Tiago. Thank you, Matt. Thanks for having me. Yeah, I'm really excited about this conversation because the competitive dynamics, the cool stuff that's being built, this is going to be a really great one for the audience. But before we get into OutSystems specifically, I want to talk about your world as a CIO right now. I find it really interesting talking to CIOs and just how they're coping with everything.

1:25Matt Paige:It feels like every 24 hours, something new, some new AI tool drops and everybody's losing their mind. But how are you processing the pace of change and specifically like you personally, but then within the organization as well? So as a joke, I tend to say that I'm probably not going to be able to know how to talk about any other topic at a point in time. Because I spend probably every single hour that I'm awake thinking, dealing, processing, managing this. I am seasoned enough to have lived through the client server, web, cloud, mobile, IoT, et cetera, big trends. And I haven't, I've never seen anything like AI in terms of the impact it can do in terms of digital transformation.

2:16So I think that on one end, it's about how do you inject in this case of today agents into your landscape and your company and you embed them as part of your workforce and your processes and your operations. And at the same time, the pressure and the speed to do that at high velocity due to the all the AI coding components and features that are coming on top of the platforms. So we live in a very competitive market right now. I think in general across every industry, but particularly in this tech industry. And the pressure to accelerate the path towards the outcomes you want to achieve is getting higher every single day by the hour, I would say.

3:05So it's incredibly overwhelming. At the same time, there's a lot of human beings in the process that you need to mobilize and make sure that they are also on the same pace. So it's a crazy but fun world at the same time. But I would say that in a very concrete factor, it's I think like any other company, we're looking at the big company bets and literally organizing ourselves to be able to go and attack those with all our agentic capacity that we have in hands.

3:39Matt Paige:I totally agree with you. I do feel like this moment is unlike anything else. What do you attribute that to? Is it the agentic nature and just the pace of change? What is it for you that feels like this is bigger than any other moment we've had before? So I think that on one hand, it's definitely the pace of change that things are happening. So you are having improvements in all this technology over the period of time of a couple of months, three months, where usually it used to be years before. and you get to a point where it's already like drastic change. And then a few weeks after something else comes along and the change is even bigger.

4:24And this is happening every day. Like you will start your day, you open up all the news outlets you have in hand and all of a sudden something drops and it's like a bomb.

4:35Matt Paige:Quick break in the pod. If you're listening to this podcast, chances are you've been thinking about how to actually use AI inside your business. And that's exactly why we built the AI Opportunity Finder. It's a free tool that helps you uncover high impact, tailored AI use cases based on your business, your goals, your pain points, and your industry. No fluff, no generic use cases, just real ideas that fit your business and the ranked by ROI potential. It takes about three minutes to run and it's like having your own personal AI strategist for free. If you wanna try it for free, check out the link in the show notes or go to hatchworks.com backslash AI-opportunity-finder.

5:14Yeah, so I think that's definitely one of the facts. The other thing is that when AI showed up about three years ago after ChatGPT came along, I think that we started seeing that it was going to be dramatic. And then you started seeing the impact that is actually capable of generating. And it has just even improved since then over and over again. For any company that needs to transform, and I don't know any that doesn't, at least not within my reach, I think that pressure is just becomes even bigger because now leadership knows that there are tools to be able to address that change. And so the pressure is very high.

5:58People need to learn fast and they need to bring these things and they need to start injecting and infusing them in the work that they do. I think that creates a lot of pressure. Right there.

6:09Matt Paige:Because I agree with you. I think all leadership teams, unless they're literally living under a rock, they feel this moment. They know the gravity of this moment. But getting your teams, especially when you're in a massive organization, getting them to both embrace it, not be scared by it, and also not feel inhibited by it. I think we were chatting earlier. You mentioned something about your job is not to put on the brakes. Like, how do you think about enabling your organization? Because it's as much change management, I feel, as technology related. Yes, absolutely. And I think that a good part of my time I spend on doing that just even today, like I spent a number of time talking to people, trying to make them understand what the end goal is, where we want, what we want to achieve.

6:59and so that they understand that the role is actually not to put the brakes on and to release, but being able to learn how to manage the risk. Putting the brakes or saying, no, it's the easiest solution people have in their hands and where they have the biggest level of control. Opening up and knowing how to manage the risk is the hard part.

7:20Matt Paige:And this may tease where we're going in a bit, but building is different now, right? you have the agentic nature, the probabilistic nature of AI. So A, anybody can really build anything now, which doesn't necessarily mean you should, but there's also the agentic nature of stuff to where automation just completely got turned on its head in terms of what's possible. But how do you think about building things in today's age, focusing around the problem and determining, okay, should this be built? Do you have like a pattern or a framework when you're thinking about, should something be built, whether it's automating some small internal process or strategy for an entire massive company?

8:04Yeah. So I think there's a number of vectors that you can address. One of them that is very much linked to the way, let's say computer science and led more like enterprise systems have been managed over years, which is you You have a problem and sometimes you always think that your job is to layer the foundations, work, like try to fix the mess, sometimes re-architect and so on. What the agentic and AI is bringing on top of this today is that actually you shouldn't actually spend time doing that. You just bypass the mess and get the agents to fix and fill in the gap. So that's a huge thing because how many people have you talked to that have big legacy modernizations or when there's a merge, like merging two different systems from two different companies and things like that.

8:57Today, it just like bypass all that mess and you're able to build and layer something on top that helps you address that challenge with whatever that challenge is. And you're trying to solve for a specific problem. So that's definitely one. And then the second one is everything that is not agentic these days is literally legacy work. And it is outdated. It's not the way you solve problems today anymore. You have to, it's a completely different and shift in mindset on how to approach the problems you have in hand.

9:33Matt Paige:It's such a good point too, because I think that's where a lot of people get stuck is in their brain. they have, okay, this is how I do it today. I'm going to automate what I do today. But with so many people we talk to organizations, leaders and whatnot, you almost need to take a step back because you can do things now differently and more efficiently than you have before. When you think of tasks and processes and things like that, the beauty is too, like you can actually leverage AI to help you work through that process, which I think is a meta concept that a lot of people don't do. I know like you do that.

10:08Matt Paige:I'm curious, what's your approach when like having AI is like that thought partner going through the process and whatnot? That's an amazing point you touched on because I think that sometimes people lose a lot of time trying to come up with use cases. Just literally think about a problem that you need to solve. And I'm talking about in the company, it could be you have to reduce costs on a certain operations or you want to optimize your funnel conversion or whatever, just any problem like that. And you literally, you can use AI to help you with that. It literally sit in the problem, give it all the context and then ask, help me out, figuring here a solution, an agentic solution to go and solve this.

10:56And you'll be amazed at how it links and connects and gives you a straightforward answer that you probably were like not even at the beginning thinking about. And then you just iterate on top of that and add more and more layers. And then before you end up with a fully automated solution that has different types of agents in between that is doing a work that was never done before. And it's completely filling in the gaps between the different steps of the process. Like, for example, sales is a good example. We always knew that being coming prepared for a meeting is a good thing to do. But a lot of the time, the research work to prepare for that meeting could be very hard and deep.

11:41And a lot of people do that because they like to be prepared. But they spend hours and hours doing that, going everywhere, trying to find out more about the people. Today, you have that just like that as a snap. So you can build a team of agents that literally collect that briefing for you, both with internal, external, public information and whatnot. And build a deep research about that account, the people you're going to talk to, correlate their challenges with what the products we sell, and even gives you tips, hints and tips about how to pitch that during the meeting. now this becomes available we can even try to understand when your meeting is going to be based by looking at your calendar and bringing this to your inbox maybe a day or a couple of days before a week before but you can also get it on demand so imagine and when you get it on demand it literally takes minutes something yeah that usually only executives had the luxury of having something like that because they would ask their people hey like i'm about to meet with this customer give me a briefing about it and there'll be one two three four five people collecting data building a slide deck for that meeting so when you multiply the hours of those five people and so on so do you have instant time saved but when you multiply that for every single customer facing person you're creating a power that was unbelievable before and this is just a starting point never mind And then helping you out through the meeting, giving you feedback after the meeting of how you manage the objections.

13:19Helping to follow up after the meeting with emails and presentations and decks that you can send. And nudging you, hey, it's time to call Matt again. You left this at this point. And I looked through your email and there's no response. So you might want to go back and tell. But just like literally it is able to put that person only focused on the time they spent in human interaction. literally and just obviously and prepare and read whatever it's prepared for that and i think that's a huge power that never existed before totally yeah for anybody listening just like

13:55Matt Paige:go back a couple minutes and just turn it on voice mode and let tiago talk to your ai to set it up and give it the context but like it's funny you say that because we built that same thing um at hatchworks ai because it's it's so beautiful because once you have the context from the meetings, through the sales process, through delivery, just so powerful. And I think the core thing is it comes down to context. And if you are, you know, identifying where are the opportunity areas, the one bit of advice I would give is the more context, the better. Yeah, absolutely. AI thrives on context. It's like the single most critical believer.

14:29Matt Paige:If you provide sufficient context, there's very few things you can't do with AI. Yeah. But for any thoughts on that part before we jump into the. I don't know if that's real. What I was just going to say is if you stop for a moment and think, let's say, five years back to make it even further. What we would do if we wanted to optimize that work was we would add more fields in whatever in your Salesforce. And we would put more buttons for you to press on. And we would try to collect data so we know what was happening and what you were doing. And you would have to do a lot of manual work that over the last years, and particularly with the pressure of public companies, and announcing results and so on.

15:12I'm sure that if you talk to a lot of salespeople, they would tell you they started spending more time in the office preparing for those forecasts and so on than they actually spent in front of the customer. So we're trying to bring the days back of don't worry about collecting all the data. We'll do that for you. And by the way, we didn't help you be prepared. Just go out and spend time with them and just give us a note. You can even talk to this voice member and we'll take care of it. for you. So it's just, it's having, it's more than having an assistant. It's like having a coach, having a sales, having a boss that's with you all the time, giving you direction what to do.

15:50It's unbelievable.

15:52Matt Paige:It's funny too, how quickly the base, the new baseline's resetting. This is just becoming normal. That, that constantly amazes me. Heck, I saw the other day that Mark Zuckerberg is trying to build an agent to effectively do his job, which I think is just That's really interesting at that level. But let's shift here. I'm really curious. So for, you're about to launch something really significant. And by the time the airing of this podcast, it will be live in markets. But for people who aren't familiar with OutSystems, set the stage. What do you do? What are you building? How are you thinking about this strategically?

16:29Matt Paige:Because I think for a lot of our listeners, how you're approaching this strategic transformation within your own company. I think it's just a super interesting topic with all the competitiveness out there in the market right now. We've been creating a platform for 20 years now where it started with the ability to integrate with every system in your company. It then brought a concept. So it was invented or created after we were already out there. I think it was between our founder conversations, between our founder and one industry analyst that came with the concept of low code. but it was in literally created a very visual and graphical representation of the application or the software that you were building and then it generated deterministically codes behind it standard codes and so that we've been operating on these we created massive opportunities and massive value in a lot of our customers with the ability to bring ideas to life really fast with factors, multiplication factors of multiple factors of four, five, sometimes 10 times faster than before or than using other technologies.

17:40And people were amazed. I can give you so many quotes like, oh, like without systems, the business becomes the bottleneck because we deliver so fast that they can't keep up with. So things like that. So we've been doing this for a while. And obviously, as AI came along and we started the first ideas about Vibe coding, it became clear that we would bring the combination of the two together. So by the time this goes live, we would already have announced what we call OutSystems Agentix Systems Engineering, which basically is through what we call OutSystems Mentor, which literally allows you, you don't even have to go now and drag and drop components and configure the parameters on the side.

18:27You literally just talk your way through and tell the platform what you want to build. There's a big beauty behind this that is you literally can generate the application using, let's say, vibe coding. But there is some magic behind, which is from the application that you generated to the codes that actually runs, that part is deterministic. which means that the only part that is not deterministic is the application artifacts and components, not artifacts, the real components. And that is very precious because behind the compiler is deterministic code generation with all the security practices and all the right way of doing things.

19:15So a lot of that is protected through the OutSystems model language. But then the ability, and we're talking about either having mentor as part of the, your ID, your development environment, you can also have use CloudCode as the interface towards that, which is another big place where people are today. And you, the beauty behind OutSystems is that in one, in the same platform, you were able to have your infrastructure. So all your DevOps components automated. This is part of our OutSystems developer cloud. We have the concept of one click publish, which means you don't have to worry about where you're going to put the application, just literally press and it promoted through the different environments that you have.

20:07then you have this ability to create software that integrates with all of the systems in your company. So if you're an existing company and you have a lot of information and systems, you literally, you layer data fabric on top of that and you can see all the tables, all the fields, all the data in your existing systems that you can use operationally in an application that you build. Then you're able to build different UIs on top of that or agentic skills. And then you can deploy that as part of a user interface in a specific channel like Slack or WhatsApp or whatever. Or you can just deploy it as an API slash MCP server that you can embed somewhere else.

20:53And you do that within the same platform. Everything. from the so one thing that people usually are not aware of is that when you build an agent before you build an agent that does a number of things you have to give the agent tools those tools are components those tools are components that allows you to access data to process information or to do things on your own on your behalf so you have to build those tools first and And then you build the agent workflow. Without systems, you do all of that in the same platform. And now with Mentor, you literally, you can just, let's say, literally vibe code your way through it.

21:38So it's massive. It brings an acceleration that I don't think was ever seen before with the ability to deploy your application straight away and with access to all the systems you have, including your single sign-on in your enterprise and everything. And that's like from the get go, as soon as you start building these things. That that's funny. That's an amazing thing to achieve, to be honest.

22:05Matt Paige:It's funny. It's like y 'all were the OG vibe coders before vibe coder vibe coding was a terminology in a sense. It's cool to see this like evolution taking place. But the point I think it's really interesting is you get to work within the parameters of out systems and everything already being connected. so it makes it a lot easier to actually build without systems. And now you have this component to where you can work in this new modality, this new way of conversationally building something. And the thing that I think is really powerful that a lot of people miss when they're building systems, workflows, apps, integrations, it's marrying the probabilistic nature of software and AI with the deterministic nature of software.

22:52Matt Paige:those two things married together become powerful and I think that's what a lot of people miss because there's things that you still need to work in a deterministic fashion and when you enable that with the probabilistic nature of AI it becomes really dynamic I feel like it's the best the best word for it yeah I think that by the way we did talk about being able to show something and actually the whole concept I was telling you about is available here on a slide this is exactly what I was talking about. So this is the whole platform. At the bottom, you have your kind of infrastructure component.

23:29So your governance and deployment. At the core is the OutSystems model language, OML as we call it. Your ability to integrate with everything and that's represented here on the side between external systems, data sources, or even the different LLMs that are available out there and then you can design EUX UI interfaces or workflows and business logics or just integrations and data or skills and behaviors for agents which you then deploy here at the top as applications through channels or programmatic endpoints like I was mentioning. I think this is a good graphical representation of what I was talking about.

24:11This is incredibly powerful because you never leave the same place you're always in the same platform and it's literally like you like spread the different things you're doing combine them all together and just publish and you're good to go that's cool and i'm curious how you're thinking

24:30Matt Paige:and you're the cio so maybe you're not the best person to talk to but like i'm always curious about the user experience and the design because users today about systems are used to the kind of drag and drop modality. How do you think about marrying the speak something into existence with the existing out system system? And then I feel like the other part is now educating and helping your current customer base make that transition. And it's almost like giving them that like light bulb moment of, wow, I can do this in a way more efficient way now. One of the things that we do a lot is we like to grab a challenge or problem and literally try to build something within minutes or hours to show the customer the power of the platform.

25:25And it's not so much how quickly you build like a simple application that doesn't touch anything it's about building something that connects with different applications in the background that goes and fetches customer data and so on and so forth so that you can see the power of all those things combined together but when you like I actually take a lot of spend quite some time talking to customers and one of the things that I do because it's like a CIO to CIO of thing is to actually bring sometimes my team and show them exactly what we build, show the internals of the product and how it looks like, and be able to demonstrate how easy those things are.

26:11Yeah, totally.