Join RedMonk Co-founder James Governor and Alois Reitbauer, Chief Technology Strategist at Dynatrace, on a fitness journey exploring how organizations can thoughtfully adopt AI without losing the human skills and judgment that make it effective. For organizations looking to get “AI fit” without pulling a muscle in the process many need to resist the urge to sprint toward full automation. Like any good training program, success requires establishing a baseline, setting clear KPIs, and building a continuous feedback loop—the very principles that underpin good observability practice. This conversation draws a throughline between observability and self-awareness, arguing that you can’t optimize what you don’t understand, whether that’s a distributed system or your own body.
This RedMonk video is sponsored by Dynatrace.
Links
Transcript
James Governor (00:00)
Morning Matt, how you doing? Yeah, here we are again. God, it’s so beautiful today.
Alois Reitbauer (00:02)
See you.
James Governor (00:10)
So, hey, Alois good to see you. This is Matt. Good morning.
Alois Reitbauer (00:12)
Hey James!
Hey good morning! Welcome to El Sanctuary, the spot. The Sanctuary!
James Governor (00:19)
Sanctuary,
blue sky, well, grey sky, depends on the day. Here we are. We’re here to talk about observability. Yes. We’re here to talk about AI. I’m James Govner. This is Alois Reitbauer from Dynatrace. I’m from RedMonk. So anyway, we thought that we should talk about the muscle you need to work in order to get fit in order to be able to use AI in your organization.
Alois Reitbauer (00:46)
Plus the muscle to keep as you’re using AI. to not lose your muscle by not working out and letting AI do everything.
James Governor (00:51)
Definitely want to keep your muscle tone. You don’t want to let AI do all the work for you. So basically we’ve got an idea of structure and progression. So you have to start somewhere and you don’t really want to jump in at the deep end. You don’t immediately start doing pull-ups. You have to warm up, you have to learn new skills, you have to train new muscles. And from an observability perspective, tell me a bit about the muscle that you need to train when it comes to AI.
Alois Reitbauer (01:20)
When it needs to come to AI, first of all, you need to understand where you are today. Like you need to more or less understand your estate, how good are you, and also what do you want to get actually out of it. I mean, here we have a very clear goal set to us, which most of the AI projects, people just start to do something with the AI. That would be us coming here. Let’s jump on and do some workout. What we decided to do, we have a clear goal there. We need to see where we are today. That’s, assume where Matt’s going to help us. Okay. Help us progress. A baseline of fitness. Baseline is always good.
James Governor (01:44)
A baseline of fitness.
Alois Reitbauer (01:49)
Then have a clear goal, set your KPIs, measure as you go along those KPIs, take in the data, take in that feedback, and then see how far you can actually go because you might realize it actually is going to take you longer. Maybe we want to do 20 pull-ups today.
James Governor (02:03)
that
feedback notion because if you think about observability, what are we actually doing? We’re trying to understand the behavior of the system. And it’s always about a feedback loop. So for example, think, you when I when I one of the things I’ve learned is I’ve been on my training journey is that the the better you train is based on how well you know each so here, if I’m warming up, I want to feel my hammies, you’ve got a feedback loop going. If I’m going to be on the bar, when I first started,
I didn’t have a feedback loop between my lats and the rest of work that I was doing. And I think it’s the understanding and the observability of the system that enables you to actually perform more effectively as a whole.
Alois Reitbauer (02:46)
You know what it was for me? have my trainers always remind me to breathe. Breathing is very important. But you know when you get really tense, you forget to breathe and you start to feel and totally forget about it? So like also getting all the signals in that you actually need to get in there and checking whether those signals are missing and trying to figure out why they’re not there. mean, obviously breathing is one of them, feeling your muscles being in a good shape, like have a full understanding of your system, which happens to be…
James Governor (02:50)
Breathing is important.
Alois Reitbauer (03:14)
our body here and then use it for our goal which is then the business observability and see how far we get there. But if we don’t understand the body it doesn’t make sense to measure how many pull-ups we’re doing anyways because we’re not even knowing what state the system is in and what it’s actually capable of.
James Governor (03:27)
I think that’s right. mean, I think the organization needs to like breathe and relax and not feel too stressed about the AI journey. You can start small, you learn some new skills, perhaps your developers start to use some of the coding tools for code assist. And really, to start with, you know, let’s just start with, can I tab this? I made a prompt. Can I accept this change? You know, we start about, you know, make sure that it’s not all fully automated, keep the human in the loop. And I think
relax into it is really important.
Alois Reitbauer (03:58)
So
that’s how we look at it. It’s almost three phases. The first is more or less augment. You are doing the task and you’re augmenting the task by using some AI. You’re still in control. You’re doing the entire task. The second one would be assist. You’re outsourcing one smaller task to AI and let it take care of it. Like look up my run books on that issue. Give me some ideas what could go wrong. Show me similar issues. You delegate the task and step three is then really to outsource.
the entire task to outsource the goal. People should take those steps. They see the end goal, but the other steps are important as well because as you progress through those steps, you learn what really matters, where you get most of the value, and surprisingly, might just be those little changes that get you way more of optimization and way more out of it than you expected.
James Governor (04:46)
Okay, yeah, you don’t always know the thing that’s going to make the improvement. So I think one of the things that you like to talk about in conversations that we’ve had before is being proactive. And one of the possible benefits of why an organization should be using AI is, yeah, we talked about you’ve got to have the baseline, but in order to start automating so you’re not swamped, like there are so many events generated today by so many systems, using AI so that you can actually reduce the false positives.
Understand the baseline so that you say that enables you to be more proactive
Alois Reitbauer (05:20)
Yeah, so proactiveness interestingly starts with managing reactiveness really well. That sounds a bit counterintuitive, but that’s what we have really seen how it works for people. you might think everybody would go into prevention first and then into remediation. What we however see happening in the real world, people are so stuck in the remediation, it takes up all of their time that they don’t get into prevention, actually kind of related to working out as well. You’re so stuck living your life, you forgot preventing taking care.
that your body still will feel good in 10, 20 years from now. So you usually help them to offload what you can on this more or less reactive side, getting most of this automated, which you can usually do very well. A lot of these are very well understood procedures. And then you can move into preventive. And this is, again, a feedback loop that helps you so that you go down on the reactive part.
you have more time for preventive. As you do more preventive, you have to do less reactive. So you just have to jump in somewhere on that journey. And usually it is interestingly the reactive part and then moving to preventive.
James Governor (06:20)
Let’s do that journey a bit more. I don’t want to pull any muscles. I want to be preventive about that. I want to understand my baseline. I want to work the muscles so I don’t pull a muscle and I don’t get injured. So yeah, that’s what I’m talking about. AI?
Alois Reitbauer (06:34)
Observervability
James Governor (06:36)
And don’t forget to check out like Dyntrace.com see what they’ve got going on from an AI and observability perspective. They sponsored this video. I’ve been having a great week with Dynatrace. Thanks so much, Alois, and we’ll see you soon.
Alois Reitbauer (06:48)
Thank you, James.
