From Network Engineer to Applied AI: Ryan Booth on the Career Path Nobody Maps Out
Transcript
All right, you're listening to or watching because we'll publish this to the YouTube's as well, another episode of the CTO Advisor podcast. I have with me Ryan Booth, principal owner and consulting consultant with Blue Ridge Consulting. Ryan, welcome to the show. >> Oh, thanks, Keith. I appreciate it. Thanks for Thanks for having me on. Thanks for the chat. >> So, this has been a year in the making. I reached out to you back in I think June of 2025 and say, "Hey, Ryan, we should get on the podcast and talk about kind of the journey of going from a infrastructure engineer architect to AI I don't know.
What do you want to call it? AI developer? What are you? >> Yeah, it's the the term that I'm starting to see come up in the industry right now is an applied AI engineer or applied AI director. Those types of roles people are recognizing that you're not necessarily building the AI, but you're figuring out how to apply it in the real world or out there for people to utilize. >> And I I don't think there's any shame in being a AI rapper.
I think What is it? Cursor just signed a $60 billion deal as a result of being a AI rapper. So, making it practical. Both of us are users and soon customers of Cursor and we find it a tremendously valuable uh tool. We'll get to that in a second. But, I'm really curious because a lot of CTOs, CIOs that listen to this show are kind of list watching their existing human resources and figuring out, how do I build Ryans? How do I get from a network engineer, a systems engineer, architect and to a AI uh applied design engineer, or whatever the title is, but how do I transfer the skill set uh which is valuable to this new domain.
So, let's first start with background. How did How did you start your the first part of your career? Walk us through that. >> Yeah, yeah. Um I grew up here in Amarillo, Texas. Um kind of middle of nowhere, but we're big enough to, you know, be on the map. Um and I I kind of fell in love with technology, and especially networking, in college. Um classical um networking person going through an associate's degree. I went through the Cisco Networking Academy.
Um and those first few classes that just absolutely beat the hell out of me in terms of learning and knowledge and all of that, I I fell in love with it. And that kind of started and pushed my career um up through networking and networking infrastructure for about 10 12 years. But then, um as SDN started coming in and network automation and DevOps started picking up, that was when I was kind of hunkering in and getting my CCIE. And so, I kind of had to sit back and just observe a lot of that.
Um and then towards the end of it, 2015, 2016 is when I transitioned into network automation. And it started out as kind of a gateway, and that's kind of the way I recommend it for everybody um that's that's in the current skill set now, is use it as a gateway. And I started out providing network expertise while software developers helped me learn my software development shops. And um that got accelerated when I moved into the startup that did network automation, Apstra. And um I spent the most of my time there just getting clobbered by the engineering teams out of Silicon Valley that I worked with.
um, basically them teaching me the ropes of, um, software development, how to write real code, how to be, um, you know, how to build enterprise-grade solutions, how to maintain them. It's not just about making it done right the first time, but you got to iterate on that hundreds and hundreds and hundreds of times. Um, and I think that's what makes a true professional in the space. >> Pause there at the Apstra part of this. I've worked with you when you were at Apstra, the CTO advisor, the before uh, Juniper was acquired by HPE.
Ap- Apstra and the Juniper folks were, uh, customers of ours. " >> Yes. >> Which has, obviously, a bunch of parallels to AI. The challenges are the same. So, I want to make sure we don't leave this part of your career without sticking a a pin in it. >> Yeah. >> There is a awful lot of parallelism between I want to identify this this like transferable skill in people. Let's jump ahead really far and juxtapose that more traditional vendor side and practitioner side career to what you're doing today.
Cuz what you do today on surface looks very different than what you did 2 3 years ago. Uh-huh. How are you using and working with AI today? >> Yeah, so I've I've basically AI has replaced my software developer skill set. Um, I very rarely manually write code anymore. Um, for the most part, I have that process in what I'm working on, um, nailed down to build out applications or test stuff or explore different areas. And so it is more of a less of a software engineer typing on the keyboard and more of changing hats all the time trying to get the best performance out of my AIs and my workflows and of course building on them.
Um but with with all that background I come from the infrastructure side. I come from the customer side. It gives me a real deep understanding of of what the customers are wanting, what they face in the trenches. Um and and a lot of times that that doesn't make the transition into the vendor space to understand how do we improve these things. Um and so that's what I'm trying to do is I I all of those lessons learned and all of those deep fights that I went through in the network automation and the DevOps days teaching executives what matters, teaching teams how how to to dig into the problems and solve them, how to build trust in the system and in your colleagues, and learn in a very vulnerable environment.
Um those were what I just basically took and put right into place with AI and just started running full force. Um >> So, there's kind of this I know executives are looking and they're asking the the question they would ask me in in this in if we were live in front of them, they'd stop us and say, "Wait, what the hell? We just jumped from like a network automation expert to a AI expert and before we hit record, you know, you're building a virtual advisor for your wife's business.
Like how do we how do we get there? Where's the in between? " There's no question to to me. I've seen like your software development skill like when you were developing software. So you understand I I a couple of different things. Obviously, you understand development workflows. That's a given. What you also understood had to understand was this integration point between the two products or the two disciplines. The network you you were a network expert one CCIE level if not equivalent CCIE well not if not a CCIE equivalent level of expertise, but also you had this soft business skill that I think uh executives need to identify.
Let's Let's drill into that. Talk to me about the combination of skills of being kind of the architect that understands both products and the architect that needs to integrate the product to solve what problem? Like what problem were you solving when you integrated Apstra and Terraform? >> Yeah, yeah. No, and that's that's a that's a that's a great point to bring up. Um and I'm going to come from the reverse side of this. I'm going to come from the trenches and I'm going to come from that guy whose first tech job was level one support Scratch that.
Not even level one support. I got to answer voicemails for level one support and translate them into tickets. From this side, you grow and you learn to build the tools that you're learning. You learn how to configure a router or switch or you learn how to set up um a Kubernetes infrastructure What whatever it may be, right an application. But as you grow into a mid and senior level, you command that area and your work pretty confidently, but you don't do well working across teams.
And when you move more from the senior level into a more staff level, um which is you see um that more common in the vendor side than I do see staff level engineers anyways. But once you get there, the key step for your staff engineers is being able to dynamically control a project and a set of goals across all parts of the infrastructure. And not just the software team, but the app team, the business team, and all those people. You you got to pull it together.
Even though again there might be a product manager there, you have to give technical direction as a staff engineer. That's kind of what's expected of you. And so I think that's that's where a lot of value comes is because that's I've usually stepped into organizations, especially from my um early days, that needed that leadership, that needed that guidance. It wasn't the technology that failed them. It was the lack of skill set and knowledge and direction of how to use the technology that failed them.
And so I'd have to come in and clean that up. And so when you do that, you learn that you spend more meetings with business people discussing business things than you do clickety-clack at a keyboard configuring something. And so that's I I've kind of built those two skill sets up side by side um because I know I can build, you know, whatever I want to build, but I eventually have to either sell or market it. And so I need those others with me.
>> So, we've established archetype. The archetype is the curious engineer who is willing to put in the work, become a CCNA, start down the CCNA CCNE uh I'm sorry, CCIE route. I mean, it doesn't matter if you achieve it or not, but this is the person that goes down that route. Uh they establish the pattern of being the one that gets stuff done. They're you're the physical management station of uh ideal a a genetic AI process. I give it a problem, it's going to solve the problem.
This is the archetype of people that's on my staff. This is who you are as a engineer. Now, let's talk about the evolution because that's the part that I think we some sometimes get stuck on. Yes, you've proven your career over and over again that you can you can reinvent yourself going from a network engineer to network automation, from network automation to business process, from business process now AI. Let's talk some level of detail of that AI transformation. What did it take for you technically to go from someone who knew how to develop software to somebody who understood agentic AI.
Like you I I've I've read some of your stuff. You absolutely understand agentic AI. What Cuz for me looking on the outside in, it was Ryan Booth, the network automation specialist who transferred into DevOps to somebody who now, wait, this guy's building agentic AI workflows. What What the hell, man? How did that happen? >> Yeah. Um I have to start with where it failed to to to cover this one. And um I'm going to go all the way back to when we first met.
And that was Network Field Day 12. And we went to And if you weren't in this one, I apologize. I I should have looked this up first. But we went to um listen to a machine learning presentation that was in private held by was Extreme Networks at the time. Yes. Um a few known faces were at Extreme and we went to a Network Field Day there. And they just layered on the machine learning discussion um to the whole delegate panel.
And all the normal actors back then were in that panel, and we were all just equally mind mush on what they were discussing. And if you go back and listen to to presentation right now, it's pretty much the same stuff we were talking about 2 or 3 years ago, um but more in general. And back then it clicked and it was like, I need to get into this. I need to pay attention to it, but I never really sunk my teeth into it.
Come 2015, 2016, I'm really deep in network automation. I'm really trying to push a number of different areas and I had people at startups to be like, hey, we're checking out this network automation stuff. You should come jump in and play with it with us. And I was like, I mean, not network automation, but machine learning. And I was like, it's too abstract for me. I've I've tried to learn it two or three times. I can't get into it.
I can't find a practical use for it in my job. And I skipped out on those, you know, those few startups then, but it wasn't until um the the first thing that landed was the the art generation stuff. >> Hm. >> And I saw it and was like, whoa, this is cool. And I I've been painting. I've been doing art my entire life. It's always kind of a side hobby. But that was my in. I can go on Saturday mornings when everybody's asleep, I can pull up Google Colab notebooks and learn how to generate AI art.
That is what allowed me to sink my teeth into it. I played around with it for 6, 7 months and then ChatGPT landed and then it just became a whirlwind. And but I just kind of kept that momentum of curiosity. And once we crossed the bridge with ChatGPT and we we could see that it could make intelligent decisions and it could actually put code together and router configs together, it was like, oh yeah, this is it. And so that's 5 years, 5 years in the making.
Um, you know, I want to consider myself a professional. Um, I wouldn't even consider myself a machine learning engineer because those guys are way, way better at that job than I am. But you put all the same smart people in a room from all these different areas, and I'm going to connect the dots for them. >> So, I think that's the key pattern to to recognize here is that you have to give your team something that they're curious about. >> Yes.
>> For you, it was uh AI art that led into you to eventually take your Agentic AI capabilities and marry them to uh application development. You've built some really cool stuff. I'll put some links to the some of the things that you build to now advising organizations on their AI strategy. I was to bring you in, and you were to help assess me on my staffing needs. What do you think the top two or three priorities when it comes to helping build this muscle internally?
What do you think that will look like? >> Yeah, that I think that it you come into um any process like this. I think you really got to sit back on and look at the workflows, look at the how how the normal business or how the normal teams work through their operations. " never really flows well. Um especially when you have a ton of people in the engineering space and in your engineering team who just don't have the confidence and the skills yet.
Um but you come in and you find those workflows, and traditional traditional workflow management product management type stuff as you find the bottlenecks, and you start squashing them, and you start working on them. Bottlenecks don't necessarily have to be a person. They could they could be something that can AI can pick up and make happen for you. Or it kick the can down the road a little bit better um to to help your team work better. I I think right now we're we're really hitting a point where everybody is is overloaded with work, um, has too much going on, and we have to make more happen.
And I think instead of us getting deep heads down into just getting it or making it happen, we need to start looking at these workflows and improving them. Um, make the AI work for us so we don't have to as much. >> Yeah, right. I love that you went somewhere completely different with the concept than I uh, than I initially presented. " And you said, "Slow up, Keith. Let's talk about practical, pragmatic AI. What can this tool do for me today?
Like, I have these processes, I have these bottlenecks, I have these tickets, I have the the things that I need to either summarize, I need to help my people. I'll lead to the HPE story around uh, the town of Vail of how they didn't use AI to replace people, but to be able to go back and summarize microfiche data. Like, that was transformative for their business, and not necessarily adopting AI to replace people, or having their people learn AI. They had their people be able to leverage the tool in a way that's practical.
All right, that's all the time we have for this episode. We're going to do some really amazing experimentation on LinkedIn. We're going I'm going to come out with a problem from a problem statement from one of the buyer rooms, and you're going to help me kind of walk through it from a technical perspective. We're going to do this as a live, uh, async discussion on LinkedIn. You want to follow this conversation, follow either me or Ryan on LinkedIn. com. Ryan, where can they find you?
>> Oh, man, LinkedIn is probably the best place to find me. Um I should be under my name Ryan Booth. Um I'm also on Substack. I have my blog going there. Um it's actually under Abstract Ryan. Um you'll find me, but you'll you'll find a link there. Also my portfolio page, this is where I do a lot of exploratory work around advertising and AI um for personal branding and stuff like that. So I have my portfolio out there that shows a lot of my work.
Um it's agentic friendly and it's interactable both by a human and an AI. If you want to see what that looks like and how it can be handled, um feedback's always welcome. >> All right, Ryan. Thanks a lot for joining the program. Uh talk to you guys next CTO Advisor podcast.
