How Platform Engineering and AI Are Redefining Who Gets to Build Infrastructure

Hey, everyone. Welcome back here to Techstrong tv. I want to introduce you to my friend Pavlo Barron.

Palo's been on with us before. He is the co-founder and CEO of Platform Engineering Labs, Pavlo. It's great to see you.

How are you? I'm doing great. Thanks for having me again.

Always a pleasure to be here. Absolutely. We love having you on Pavlo.

I, I mentioned your co-founder, CEO over at Platform Engineering Labs, but you know, you have a lifetime of experience in in technology and infrastructure and it tell, give people a sense of how you came to be the co-founder and CEO over at Platform Engineering Labs. Oh, well, that's, uh, a very, very interesting question. Um, uh, I mean, there's a technical side and there is a, uh, visionary side to it.

So technically my co-founder, Zach Schneider, we worked together in a previous startup, um, um, called Instana that exited to IBM 2020. And we were working on a lot of things around, uh, making a very complicated, never stopping platform 24 7, multiple petabyte in a short period of time, data ingress, making it work. And, uh, that is a special challenge.

Probably not everybody has a special challenge like that, but what you learn is that the tooling that you have at hand already, like six, seven years ago is, well, I wouldn't say outdated, but it's so rudimentary, it's so low level that you basically crave improvement and kind of use case orientation. Which leads me to the visionary part of the story, which is that him and I, we always kept talking about that since then. Um, like how can we improve that?

Because nobody cares about use cases, concrete use cases in that space. Everybody works and plumps together. Um, you know, any workflows with, with very basic tools, and it's old toil.

It's, it costs you time. There is no way you can really safely test it or even run it without having to chant and dance around it, uh, uh, for, uh, weeks. So, long story short, at some point, um, um, well, a little bit more than a year ago, we just decided to change it.

There was a few triggers, uh, in the market and we said, Hey, it's time because there's platform engineering and platform engineering is a discipline. We actually agree that this is the next evolution that is necessary. And, uh, the great thing about it that it's, it finally admits the truth that every it was running on, which is a few real good people run stuff for many, many, many other people.

And that is, this is the, this is the way to go. This is how it always worked. Now we are honest with ourselves that this is how it, how it's supposed to work.

Unfortunately, platform engineering itself is inheriting so much stuff from the past disciplines and decades of development of law level tooling that it's actually, this tooling is in the way because platform engineering itself is all about use cases. It's not about tool chains. It's not about pluming together chains out of hundreds of tools.

You need to understand that the same engineer has different hats on at different times, uh, of the day in different situations. And this definitely needs to be taken on a new level. Absolutely.

Absolutely. com, right? So I, I live this every day, right?

Talking to people, listening to people re seeing what the, the, uh, the community, uh, contributes in, in terms of, of articles and so forth. Um, but we, we had a very interesting discussion off camera that I want to bring on camera, right? Because you, you guys recently announced over at Platform Engineering Labs that you expanded your format into a platform for infrastructure builders with multi-cloud support.

We're going to get to multi-cloud support in a minute, but let's leave that to the side for a second. You said something that I think was really important, which is, in today's world with ai, we're all builders. All of a sudden I'm coding, right?

With vibe coding and stuff. I've never considered myself a coder. My writing and my discussions and my every I I build, I have my, you know, we have mu some of our a v team have music backgrounds, Grammy Award sound engineers, they're creating synthetic music.

They're building music videos. Um, our video people are building 3D animations. They weren't 3D animators, they were video editors.

So one of the great things I think about this AI is that we're all builders, we're all builders within the, uh, the IT realm though. It's really, you know, expanding, mind blowing in terms of everyone's a builder and what they're build. So when you say platform engineering labs is expanding the format into a platform for infrastructure builders, What do you mean?

So this goes back, um, uh, history wise at how IAC actually, or infrastructure as code actually evolved. It has started with the idea of, uh, declarative kind of infrastructure code that everybody can contribute to and use. Now, the problem that started appearing at some point is that this whole infrastructure has become more and more diverse, more cloud providers, more different services managed unmanaged stuff on premise versus cloud, um, kind of hybrid scenarios, your own APIs, gillions of different tools that are involved in workflows and whatnots.

And this is still, all of it is kind of infrastructure or at least needs to be taken into account when working with infrastructure. The complexity has been growing and at some point people started extending those IAC tools like Terraform with new providers, et cetera, et cetera. Now, the programming model they follow is actually designed for top-notch engineers.

And, uh, they take into account that extending the systems needs to be done by like a classic software engineer who's very well educated and, and knows their stuff. Now, this is not necessarily this kind of profile that operates in the DevOps, uh, in the broader DevOps space. Uh, they might be software engineers by heart or by education, but sometimes they're not, it doesn't mean they're not top-notch, but they're not used to this modern software engineering approach where you kinda, you know, this, the ecosystems are big and, and you need to, to do so much just to write a little bit of coders, all this testing around that, et cetera, et cetera, et cetera.

So, um, we're talking more about people who have been solving tasks by scripting things. And, uh, when the scripts do what they want, it's good enough. Now, those people didn't have access to implementing providers for this classic platforms like Terraform Plum WhatsApp.

Now, right now, as you said, we live in the age of builders. I mean, this is the AI idea or the, the, the advances in this technology had actually accelerated everybody who's creative enough into the mode of, oh yeah, I can plan together whatever I want. If I know how it's supposed to work, I don't really care about it's, you know, workings as long as the result is what I need.

And this is the similar kind of yeah, mindset. Uh, for those people who have a task at hand and need to achieve these tasks iteratively and have this result. Guess what?

This is scripting, right? This is how we always use, uh, to write all these scripts in the operation space. Now, what I'm trying to get at is, um, there is a gap between how you can do that in classic systems in, in classic platforms where you have to wait for somebody else to write a provider, a pretty complex piece of software, because the platform itself does not really abstract anything for you.

So the software that you write as a provider is very complicated. You need to wait for somebody until they really start supporting that. If you don't agree with how they did it, your only choice is to make it work yourself.

Sometimes you can't, you don't have the time, maybe you don't have the qualification. I, I don't know. Now, in this time with, with the help of ai, what we actually are expanding the platform into is what was necessary for IAC always, but now it's very simple and, um, safe.

What this means is, yes, you can write code your infrastructure code against the plugins that we already support, like the technology. We talk about the multi-cloud support. But, um, if you, if you miss something, you don't want to wait for us or our community to implement something like GitLab support for you or your Jira support, whatever, maybe we're slow, maybe we don't have the priority for that.

You just go ahead and do it yourself. We give you the tool that you actually feed your AI copilot or your AI agent with, and within a few hours you get to a working result that is also saved and completely verified against the sophisticated test suite, if you wish, which completely exclude a failure. Uh, I mean maybe it sounds bold, but you can engineer it and that's what exactly what we did.

So everybody who knows what they want, who has access to the documentation or source code of what they want, is able with support of their AI copilot to achieve, um, uh, technology support and put it into form neatly within hours. If you want to do it all by yourself without any AI agents, you still will achieve that within days. I mean, you just slow down yourself because they have to type all that, right?

What a great time to be alive. Ha, Palo, did you ever absolutely think back, think back to when you first broke into it and you first, oh God, Right? If I told you this is how you were gonna do things now and don't bother writing the code, this'll write the code for you, Would you?

Oh, yeah. I mean, I remember the first, uh, printouts of the manuals and everything. I mean, you really, you really have had, had to be absolutely creative.

And I mean, I'm really old. I come from the time where typing actually, uh, was expensive. I took typing in school, Right?

Yes, I know, but there is typing and typing, right? I mean, when you, when you go back to punch cards that Me Too. I was there.

That's what computers were when I was in school. We had punch cards. Exactly.

So, uh, there used to be a time where you even typing was, uh, too expensive for you and you need to, to be, uh, thinking with a computer together. The times are over. I mean, yes, I know that there's a lot of people who are nostalgic about all that, but for the regular developer, for the regular engineer, we need to move forward.

Moving forward means just taking away all the, uh, potential, uh, kind of stones that lie in the way to achieving the result. And what Foreman now delivers to you as an engineer is there is absolutely no difference anymore between you being able to write the declarative infrastructure code that is always up to date in our system and extending the system that will allow you to do exactly the same thing with a piece of technology that we don't support yet. So that's like just a natural next step.

Absolutely. Um, Palo, I need to pivot a little bit and talk a little bit about platform engineering labs. 'cause we're learning low on time.

People listening to this and say, wow, this sounds great. I'd like to check it out. I like what Palo and the team are doing.

What's their best way to engage with you? Oh, this is, uh, perfectly simple. I mean, we are open source.

This means that everything is on the platform. Dash, uh, engineering dash labs on GitHub. There's plenty of repositories for me itself is there, but also a growing number of plugins that we're supporting, um, for different technologies.

Um, you can check out a blog. There's also links, all kinds of links into our Discord community, into our blog. We are very active on LinkedIn, so there's plenty of content you can learn from, agree or disagree, but it's, uh, interesting enough to, uh, basically follow.

So we are completely open. We just want people to take our software and to start building without, you know, unnecessary cycles. I love it.

And the website. Yes, it's platform engineering. Platform engineering.

That's great. Palo, as always, it's a pleasure to have you on here on Tech Trunk tv. Keep up the great work.

Don't be a stranger. Come back and talk with us more Anytime. Looking forward to that.

Alrighty. Palo Barron, co-founders. CEO Platform Engineering labs.

That's platform engineering, platform engineering labs on GitHub. Uh, go check out all of the great stuff he's doing. You know, we're all builders today.

We're gonna take a break here on Tech Trunk tv. We'll be right back.