Report on AI-Augmented DevOps – DevOps Unbound Roundtable

com Roundtable brought to you by Tech strong and tricentice. My name is Cody J brownham the host of tech strong learning and we have an exciting panel for you ahead. First.

I do have just a couple of housekeeping notes to cover. Today's session is in fact being recorded So if you miss any of our discussion or you'd like to rewatch or share with a friend be on demand recording will be made available shortly after we conclude our live session today. If you have any questions, and we really hope that you do we want you to send those into US using the Q&A tab, which can be found on the right side of your screen.

That's also where you'll find the chat Tab and that's where we want you to get to know your fellow audience members, or maybe just let us know from where you are joining. And of course before we close we are giving away 425 Amazon gift card. So be sure to stick around for the duration of today's program.

So welcome everyone to devops Unbound this episode report on AI augmented devops, and I'm joined by Clint sprou director of product marketing at tricentes. Kosuke kawaguchi creator of Jenkins and co-ceo of launchable Mitch Ashley CTA CTO of Textron group and principle of tech strong research. And Mike Rothman Chief strategy officer for Textron group and GM of tech strong research.

So Clint kosuke Mitch Mike. Thank you all so much for joining me today Mitch. Do you want to go ahead and take it from here?

Fantastic. Thank you Mr. Cody.

Jay Brown. Appreciate you kicking us off there and style and a special thanks to our producers today both the linear and Cody from and so linear and Jody said she will pay for that later. I'm trying to do a good thing.

So no because I wanted to mention that this this devops and bound is a series that we do by weekly. It's it's been a long term series we've done now for a couple of years and tricentis has been a sponsor of this pretty early on and and with what's interesting about this is Trace Sanchez is a company with with texture and group has been really more of a partner and putting on the content for this and we have a little bit of a special episode today because it's both alive. It's our Round Table version of it, which we do once a month in addition to our bi-weekly recordings and then But it's also about some research that tristances asked picturing research to take on around Ai and AI augmented devops which is a topic of course, so we're going to get into that.

We have a free report and want to mention that right up top. If you go and look at the chats. It's free report.

Click on that baby. That'll download the report for you. We're gonna be talking about that but also kind of intermixing with some just some discussion about augmented devops in general and AI in the role and devops where we what where we making in rows where we might go next all that kind of thing.

So that sets up what our conversation is today. So let me Begin by asking our panelists to introduce themselves and then we'll get into our topic Mike since we're talking about research today. Why don't you start us out first introduce yourself?

Sure. I'm Mike Rothman. I am chief strategy officer strong group as well as general manager of tech strong research.

I'm the principal or you know, certainly one of the auth Is on the report, so I'll have a few things to talk about relative to that and a 30 plus year security professional. So I kind of look at it from you know, I don't want to say old guy perspective but you know, certainly a you know, experienced individual who's kind of seen a lot of the trials and tribulations of Technology Evolution over the last couple of decades and it's a really interesting time to start thinking about how we can apply a lot of these new capabilities to the problems that you know, we're dealing with obviously we'll kind of hone in on devops, but I think it's really an inside an exciting time to be in technology right now. Excellent Goods.

Okay, once you introduce yourself sure, so hello. I'm kosuke and based in San Jose and I probably best known as a guy who invented Jenkins and that was at this point 15 years ago. So I've been working on drinking for quite some time like a builty company Travis around it.

And then like as I started what it's like a automation sprawl in the software development. I feel Pride but I started thinking about okay. So what's like, what's the next year and I thought okay.

There's like all these automation producing lots of data. But okay unlike manufacturing we often like to use as like, you know, the lighthouse so very trying to go we already using that data to make the open process more effective and I hear from the you know, practitioner all the time like what why they can't get the organizational supports. I thought okay this seems like my next mission and that's what I've been doing for the last few years.

Yeah company Cola Cascade thank you for Jenkins this first tool I and I'm sure many other people used on our devops journey. So thank you. You've already had a major impact in the industry and excited about what you're doing at lunchable as well.

So thanks for joining us today. Thanks for having meant repeat guests good to be with you again introduce yourself. Yes.

Definitely everyone. My name is Clint sprou director of product marketing with triceps. I primarily focus on our devops messaging and positioning also the piano for some of our emerging products coming through on based in the Fort Worth, Texas area.

They would try says about two years been in the testing industry for a ridiculously long amount of time or several testing vendors over the years most of the you know, usual suspects and whatever reasons that I was really happy to participate in this particular round table that we're really seeing the evolution of quality. The board and we're seeing some of those things that a lot of the vendors promise like over 20 25 years ago, you're starting to you know to see that actually on you know to actually happen within the industry itself. So really happy to be here and again as I mentioned off camera just being in the presence of KK and just you know is what he's done just for devops.

It's a software development in general. So this is just great for me some I just super excited. I'll try not to be a Super Fan Boy disappear, right but you got hard to do that.

But you know, yeah, it's the nicest guy most humble. So we appreciate you thank you. So let's get into our topic just kind of refreshing an old saying Yogi Bear, you know AI everybody's talking about it, but nobody does anything about it.

Well people are doing something about it there. They're playing. Yeah to Things so let's kick off Mike.

You know, why start why this report why this topic? Why now, we hear we hear a lot about Ai, and of course, I think in everyone's mind is like when is it real? When is it Market?

Eyeball that kind of thing. So tell us a little bit about what you kick off this research with you. Yeah you bad and and I think a lot of it gets back to kind of the ethos of tech strong research, which is you know, we really want to be looking forward right we want to be thinking about where stuff is going.

There are a lot of other folks out there. That'll be happy to tell you where things have been they may do, you know kind of funky quadrants and and you know kind of wavy things and those those kinds of of reports but you know, what we don't really see is a lot of folks that are out there trying to get a sense of when right we've talked about AI for a long time, I think somebody in the chat mentioned, you know, 20 to 25 years. We've been talking about, you know, kind of a lot of these Technologies and And that's right.

So what we try to do with with text strong research is, you know, obviously leverage all of you folks, right all of our audience. It's out there and and you know, have you share your perspectives with us? Have you give a sense of where you're at relative to the evolution of a lot of these, you know main Trends and obviously Ai and and devops the intersection of AI and devops is what we studied here.

And and I think what what's interesting is what we had a couple of high level, you know, kind of findings, right and that's on the executive report of are the executive summary of the report. So we'll plug that a number of times right up there in your chat so you can grab it, you know kind of and read it right now, but Really what you know, it comes down to three, you know kind of Main High Level conclusions that we drew from the report. The first is there's a great expectation that AI is going to dramatically impact favorably right devops.

So that's one of the things that that we took away is that again, a lot of folks may not be there yet. They may be, you know kind of early on in their process either with devops or with AI but the expectation is that they are going to feel and get some, you know, very specific very tangible benefits from you know, using and leveraging AI within their devops environment. The second one is that devops and and that AI within devops, you know, kind of motions can really impact favorably a lot of the business and Technical challenges that we face with devops, especially around People right and it's like, you know, you talk to five people and you go.

Oh how many folks you know have open wrecks on your team all five of them, right? You know, how many folks have too many people nobody ever says that right? So we still have a skills Gap whether you're looking at it from security or Ops or development.

We don't have enough people right to do the things that we need to so we have to leverage the machines more effectively to help us become more efficient improve Effectiveness and the like and and that came across loud and clear, you know from the data that folks do expect Ai and devops to really have a significant impact on on what they do from devops table and the third high level thing will be music to cleanseers for sure. Is that a lot of the folks believe that AI will have the biggest impact on testing right? There's a whole mess of stuff you have to do as you scale up your environment.

Just the number of moving pieces. What are you focus on how do you make sure you have proper coverage? I mean all of these things really really kind of explode when you're scaling out, you know, a devops environment and we don't really have a choice but to leverage the machines right leverage a more analytical capability to be able to keep up and keep Pace with you know, kind of those aspects.

So at the Top Line that's what our study found is that you know, get folks are expecting big things, right? They think that they can address a member of the technical and business challenges that we're here and it will have the most significant impact on on the testing aspect of the devops Infinity Cycle. Okay jump in yeah, yeah, so when I first saw this report my neck, oh my God, I got this so this device investors and then you know, it's okay.

Well it's always helps another people saying the same thing. So that was great that at the same time. It's almost a faces I felt like this this got to be too.

So I'm looking forward to having those like a more conversation there. But yeah, I mean some sense, you know, I I kind of the trajectory these the thing they learn we think it's like okay you start to use a lot more drive a lot more computers to make our first product at some point the number become too much the for humans the interact with them directly. So it's only he's a nutshell directions.

I mean like a layer that separates you from all the glory details. It's going on directionally makes sense. Where are we today?

Well, that's that's fair things kind of gets interesting. And I was really interested in the fact that we talked just really how it all comes together, especially in the testing space when you start talking about those skills skills Gap, right? So we've gone from though they're you know pushing heavily on the programming aspect to low code no code and then how does AI affect that because what we see a lot, you know dealing with our customers and Prospects is that many of them they have these they still have these large QA organizations and they have folks that are those subject matter experts and business analysts, but then literally have the skills to really hold in on some of the areas where AI can actually really help accelerate that so that's an exciting time and exciting future and seeing how this is really going to have ai infused devops really improve.

Not only just software development in general but also accelerating quality across the board. It's an interesting time too. But there's a question that someone asked since I since I mentioned I've been doing less than prologue and expert systems going back to the 80s.

So I haven't worked in AI in my whole career. It's kind of why now, is it because of all the compute that we have available to it? I mean obviously, you know computers infinitely better than it was in the 80s by a long shot.

But it's you know, many factors one things. I threw out there is yes compute, but it's more adding more importantly if there's been huge advancements and machine learning algorithms, but I think the biggest factor is just the massive amount of data that like unstructured machine learning can leverage as well as structured. But you know now we can really we have enough data to chew on to make some interesting connections conclusions, you know relationships all that kind of thing.

What are all your perspectives? Yes, exactly that the you know, I actually thought the another area that I had in the key. So QA is clear your command there yet.

That's pretty lots of data and then so which otherwise I wouldn't be picking this other area my style but I thought the another area that was head in the devil space is actually oops more in the operations the monitoring, you know example like that those I felt like these in the area that has like a lot more adoption of a I like these people have long forgot. I mean long gave up is the idea that they're gonna be babysitting individual like components of services. It's suddenly a mass and data game.

Yes. I I feel like so the what makes it easy in the key space is like we can point to these guys like okay when the data size or the scale of things you're dealing with girls be on certain points. He's affected eye on these tools and you don't have to invent the new thing.

So awesome. Yeah, I think that you know and I'll kind of look at it from from a Dev secops piece of it too and really hit on, you know, kind of what KK just said relative to you know, the operations side of things but in you know again for the security, it was like we didn't know what we should have been looking for. Right, you know and and the first evolution of you know, any of the monitoring Technologies was really looking for specific patterns, right?

We call them signatures, you know, your AV, you know kind of your intrusion detection. I mean all of these things were based upon patterns that you need to look for, but the problem was the attackers didn't get that memo, right? They're like well if you're looking for that, maybe I'll do something else and we would have to wait to get pop by something and then you could look for it.

Right? That's obviously an exact thing. So the ability to aggregate just scads of telemetry and data and the ability to use these analytical techniques to infer relationships and patterns that we didn't know existed that we didn't know to look for that's really made all the difference in the world.

Right? So it was just it really has been a technology thing. We couldn't aggregate the data in, you know, kind of a scale enough fashion.

We didn't have the algorithms or the math to In you know kind of what these patterns were and you bring both of those together and you're just like, you know a light bulb and then you know all the math people become data scientists and their salaries go up like Forex and you know, it's it's a good time to to be in data science, but it's really been enabled by a lot of the technology Evolution that we've seen specifically around cloud. Right. Yeah, and you're also seeing that the impact of only that test data or a test data management along with security.

And so now you're looking at you know, we've all heard the you know, the term devsecops, you know developer security having developers more security. So we're really seeing the, you know, the influx of AI within those security tools as well. It's actually making that much smarter in terms of how do I manage, you know, this massive amount of data and all these different security issues whether I'm doing SAS or SAS or any other type of security testing and that becomes really big and it's I think what's really amazing me Beyond this report is when we start looking at the impact of AI from just think it just software development and some of the tools that are being developed by Debbie West Google, you know and Microsoft in terms of helping developers write code using Ai and making that much smarter and being able to actually create applications, you know from existing code bases so that all kind of dovetails into infusing that into how we develop software within a devops paradigm.

Excellent. Well, you know I think Mike maybe I'll jump to There a promise to the audience Knows Lies We are not doing it. I hate slime.

But what I will do is I'm gonna do a little bit of screen sharing. There's some sections from the report. This is the report you can download from that way up there right up there in the you know, it's yeah that's weird.

One of the other, you know, I'm doing a quick screen share and there's a couple things on here. I think we worth highlighting Mike if you could touch on or at least kick off and some conversation. Hopefully everybody see the report there on page five.

I believe it is. I've got a nice graphic here. I'm gonna zoom in a little more perspective on you it is because you know, and and again, I mean, I think that and we got one of the questions in there was you know, what specifically your folks expecting and and where's the impact and we will certainly get there right?

We've got all sorts of different cuts of the data in the report, but you know the thing and again in in terms of Really trying to look forward right and making the assumption that you know kind of the folks that have spent more time doing devops really identify themselves as mature devops organizations and we cut the data that way right. So we asked them, where are you on this, you know devops Evolution scale or you you know again just get started or you know, I forget what the four of them are. But mature devops was, you know, kind of folks, you know, basically self identifying that they have a mature stable and productive, you know program in devops environment.

So so we just wanted to take a look at that. Right just those folks that are most advanced in terms of their use of devops and really to get a sense of are you getting value? Right?

Are you finding that you leveraging AI is useful, you know to help to help really streamline and and work on your, you know, specific devops initiatives and And the data is the data, right? The data says, you know 79% of the early, you know, kind of those Advance those mature devops shops are finding value using Ai and again, you can't say well that's kind of obvious. Right but it's not because again, there's a lot of different areas where you know, we can use Advanced analytics.

We can you know leverage a lot of the data in Telemetry that we're getting there but it doesn't always translate to real value within the organization right within the process. So that's something that you know again, I thought was very interesting, you know from the perspective of getting, you know kind of access to that data. So that was kind of the first thing we wanted to highlight, you know from that expectations standpoint that the folks that have been doing this the longest identify themselves as the most mature clearly state that you know, AI does add significant value, you know within their devops environment.

So this was actually one of the shocking numbers to me because like this kind of implies. So this kind of implies that these guys are already using AI augmented devops tools and I felt like Oh, I thought like this is still a relatively You know minor or friends needs practice. So, you know is this is this reflecting in their expectations?

They hope it will be useful like when they say find it useful. It sounds like they have already used it. I'm sorry, I guess you dropped off.

So I don't know. I'm here. I'm here.

I just have it. You know, I I know so and and that's really a great, you know, kind of clarifying Point KK is that again? So these folks that again identify as mature, right?

So they're the most experienced on the devops side, you know, at least some percentage of these folks, you know have and and have started to use, you know, kind of some of these AI tools with within their process. So that's you know, kind of and again, you know, obviously there's you know so much that we can do with survey tools. So we didn't and you know kind of talk to each one of the 2600 respondents and we did have 2600 respondents in the in the survey pool.

So that was again a great response from from that perspective you can Things like significantly their statistically significant numbers and the like but but yeah, so so and it turned out that the folks that identified as mature devops were roughly 15 to 16 percent of the respondent base. So so again, that number is taken from that, you know, 15 to 16 percent of those respondents that identified as experienced. Okay.

It's interesting too and someone reached out to me with a private message. it's we're constantly living in this world of what is what's real and what's expectations and kind of blurring the lines between those two. And I think in the same way, we also get the lines get blurred between are you practicing AI?

Are you implementing AI yourself as part of your tool chain as part of your workflow as part of or is it built in Pratt into products you're using or you just being told is built into products you're using but you don't really know, you know, there's there's a whole spectrum of of this and that I imagine that you know, I wish we could get all those people back together on chat and say no which way do you right? You know, so it's got to be it's got to be in a mix of all this but I would imagine A good portion of folks said well, you know we bought this from vendor Y and they said they use AI so it must be influencing what we're doing today to you know, hey, I'm spending work my time working on it. So To me.

That's right. Actually, maybe I'm over rationalizing it you can you can That's you rationalize everything. But yeah, there you go.

No, it's gonna say you actually bring up a really good point because I remember even just you know, a few years ago when you start looking at especially it AI became this buzzword for a lot of companies to get funding or to kind of pivot, you know for their brand new product if I could just say I've got a machine learning or AI all of a sudden. Wow, you've got something new interesting and it's funny. I was looking at on page 14 of this report where we talk about how it's like, how are using AI to argument your processes and it says 49% they reduce test case maintenance with self healing now it's funny.

So you talk about separating, you know, you know fiction from you know, the actual real world. So we think about self healing if you're looking at it from a UI perspective, you know, there's some companies, you know in the past that have used that self healing Um as from a UI perspective, they kind of look at it from the standpoint of okay, is it truly AI or did you just build an algorithm to handle some of these changes to the UI? Is it truly doing self-healing?

So I think is individuals start to look at the technology really understanding how much machine learning is actually being done to be understand what's real and what is not real and I think that's huge for a lot of organizations. And I think there's so many companies that are just claiming AI but there's only in my opinion that there are select few. They're actually doing some of these things when I think about self healing I think of it as doing more than just okay, if a button change well we've been able to figure that out 30 years ago how to handle that right?

That's a very very basic use case, you know, what happens when the entire UI changes or I remove that particular component or something's happening. That's just at the top of my level. So I think understanding what is real.

Is it plays a huge huge part in how AI is really going to affect devops or let me give you guys opinion on that in terms of what you've seen out there that you know outside. You know, what's real. It was not real.

Yeah, I feel like the practitioners closer to the Grand or like that. You know, they're looking for they have this problem like that. They are trying to solve these like, you know, they're like a low maintenance United and if it's like a power by AI or algorithm or something else.

It's like it's not the I don't I feel like they don't really care all that much about if anything make these developers to stick together, but that's what they don't like this. Yeah, I watching and I feel that every day. So thank you like the healthy number of people that really but the other side.

There's also another grouping companies. He's jealousy kind of sure the practice that cross the organizations and they're open. She's like they need to get more resources that they need to drive therefore and sometimes like aligning themselves up being these like what's going on broadering the industry or the world in like a healthy management understands like the context of this and if that translates the more people money, you know.

Time for the practitioners on the ground and all the power speed. I mean to them so I feel like that's part of what's happening and sometimes when I hear like a developer shooting that these people because they're just a buzzword. I feel like guys like they're trying to help you.

Okay, so don't you have at least in front of your boss is kind of IEP. That's what marketing people are for shoot them. No, I'm sorry Clinton time is not targeting you.

Let me turn off lights. Let me turn off my camera just Anyway, I'm gonna I'm gonna change my title from marketers. Yeah.

There does seem to be a lot of you know, kind of questions just about what this whole mature thing looks like. So let me just kind of walk you through the definitions that we used for the purposes of the again self-identification. Right?

So we had four different levels are getting started. So we're doing code reviews and have improved collaboration Dev test experience and productivity. We have operationalizing devops right practicing continuous integration and delivery right with automated builds and automated testing standardizing devops.

We are practicing cicd with security testing integrated into the development life cycle. So that looks a bit like deaf secops and then we've got matured devops right Cloud native multi-cloud flexible infrastructure infrastructure as code. So those those were kind of the terms that we use to try to get folks to, you know, categorize and and classify themselves from that standpoint.

Is that perfect no, right and we need at least some level of categorization. In to try to get you know interesting ideas about where folks were on that maturity curve. Yes.

And if anybody out there has some suggestions for how we can refine that in the future. We are all ears right? Because you know, we know that anytime you're kind of surveying a bunch of folks.

There are going to be you know, some specific challenges that you have in terms of how you frame the questions and and again, we we love feedback on that front because this is a moving Target, right, you know kind of where folks believe they are versus where they are and then you get into you know, again confirmation bias and ego and a lot of other psychology things that I don't know that we're digging into you know, and in this specific episode of devops on bound but to hopefully that gives everybody a little bit more context for you know, kind of what those those definitions were. Good, I think that's that's very helpful. That folks have been Lisbon running dialogue.

If you aren't watching chat, please jump in there too about what do we might mean by mature? Especially given devops is such a journey, right? You know, it's not a destination if you will and it changes right imagine.

And again, we we're you know, being the the senior members of the the, you know, kind of panel here. I I had said something about old before and I got lit up on the chat. So I'm not gonna say that anymore like, you know, there was experience Somebody complemented by this season.

I don't know very much right. I just like that that is very flattering. So thank you on that front.

But you know, we've seen this before right we've seen a lot of these, you know technology things we started right at least Mitch and I, you know started in in Mainframe land right and and it gave way to client server and you kind of saw how long it took for folks to really Embrace that mentality and then internet architecture and three tier applications happen and It took you know time we were just getting comfortable with client server. Like, you know, the big fat clients and then we're just like, oh now we got to do this. Oh my God to say that.

Okay, and then other fat family out now Cody's laughing me. We got the explicit Family Hour my body we can have you know, but we were just getting comfortable with the fact client and now they're telling us we've got to split out, you know kind of the Thin Client and the and and you know, I guess the point is Right. We've seen this over and over again, you know, there is an unequal.

I think somebody put, you know burritos principle in there, right 80 20, you're gonna have 20% of the folks that you know, just run head long into it. You got 80% of the folks that are you know, somewhere kind of the back end of that Evolution from that standpoint and it always takes longer. I always harken back to that Bill Gates quote, right, you know, you always overestimate the amount of progress you can make in two years and you dramatically under made the amount of progress you can make in 10 years, right?

And and when we look back to what things look like 10 years ago 20 years ago 30 years ago. It's it's astonishing it really is astonishing in terms of you know, how quickly things do evolve and change and we are clearly in that process relative to devops. Yeah, I totally agree with that and I see one of the one of the challenges that I see even with a lot of our customers because we serve you know customers that are doing you know modern software development some have very complex, you know Legacy systems where there's still supporting that Mainframe those thick, you know fat clients.

They still have you know, desktop applications, you know via Citrix and whatnot that they have to support and they have all these Integrations of package applications as well. You know, whether it's sap workday service now sales Sports and so forth so that the challenge that we're seeing with them in terms of implementing a some sort of continuous integration continues delivery pipeline. Even with symbols package applications is that they still have to support a lot of the Legacy side as well as kind of move to the modern in that kind of becomes.

More of the bottleneck versus just testing alone. So just trying to make sure that you can get that level of integration get a certain amount of automation across the board. So what we kind of tell our you know prospects is we have to help them to find devops on their terms, right?

So if I try to do it the same way a AWS does it or a Facebook does it or a Silicon Valley startup? I may not be as successful. I can take some of those best practices.

And so this is where I think AI is really going to help alleviate some of that or a machine learning because you have all these various complex systems that are really slowing organizations down because they think okay. We're gonna go ahead do this without the right folks in place. Yeah.

We got all these Legacy systems over here. We still maintain this and we're trying to go Cloud trying to do Cloud native that we need to do this lift and shift and the whole nice we just gets a little bit more complicated but but like you said, That Evolution and like we've been here before we've seen this stuff since you know Mainframe fat client and you know distributed systems and so forth. Some JK, you'll say oh, yeah.

I remember we had this thing called Jenkins back then it'll still be around your career what you want to share because I want to go into you know that we did this report didn't Focus just on testing. We've got some interesting date on testing. I want to share and I'm not really a name dropper but I worked with I worked in a research organization and there's gentleman by the name of Bernardo humorman who was an expert in AI one only researchers only other name.

I know by the way is Marvin Minsky and I never met him but anyways, and I asked him it's why is this surge? Why is this researched away we hearing so much about AI why is this really kind of looking like it's becoming real when it was supposed to be year after year after year and his response was it's, you know, there's a lot of advancement but we especially see advancement in machine learning and because of the massive quantity of data, that's what machine learning needs is large and continuous feed. Data to look for this connections, you know, it doesn't know that, you know, this is a pixel move and when that happens that's that causes air.

So it just knows that's a pattern that it sees in the data through kind of unstructured learning algorithms and and Know what better area operations is one great area. Testing is another great area where we get, you know, massive amounts of data and especially in a devops world. Thank you changing KK for all the automation, you know, we generate so much data and I want to share just this one page and no charts just this one page and I promise.

Well, I think we'll limit it to this. So Mike maybe if you want to talk a little bit about kind of the automation. Side of this in some of the insights that we had.

Go down one more. I'm sorry. I did pick the wrong patient.

Nope. There we go. Yeah, there it is so we can kind of hone in or blow up on the a little bit so you there we go.

So, you know, what we wanted to do was was also dig in a little bit and and try to understand, you know, kind of where in the testing process they found or expected the best or the most significant impact of AI. So so we asked you know, a fairly straightforce wasn't you know, what types of testing are you struggling to automate right and that's you know, kind of a place where we would think that you know Ai and some of the you know other capabilities that that kind of go along with devops would you know kind of really help from that perspective and you know, we saw, you know, all you know kind of fairly close, right but functional testing and and performance testing, you know, really kind of jumping out from from that standpoint. And then we also ask the companion question and a number of folks in the chat are like what is expectation and how you're defining Ai and all that and and you know, again, there's a lot of of depth there that and only so much you can get into and and Survey Monkey, you know.

Type of environment but it was interesting in that you know it from an expectation standpoint folks did believe that functional testing was going to you know, kind of have the gray or benefit the most, you know from the the introduction of AI into you know, kind of that testing process. So, you know from that same point, that's one where both folks had a problem automating and they believed that you know AI was gonna help and then when you kind of look at the rest of them, you know, they're all kind of the same but you would think that you I would have been a little bit higher on that front. But again folks are you know kind of reflecting you know what they think at any given time so I thought that was kind of interesting date as well.

Yeah, because this has also puzzling. He said I thought today where things are used. The most is this, you know, like I behaving emulating the behavior of humans and you are testing in Mobile Brothers like that.

Do you see I mean even unit tests rank this time like what are you expecting guys? And and so yeah, this is I don't know how to explain this really very different from my perception. Yeah, yeah, I would have thought just based upon the results that performance testing would have.

Rated and much higher. I think just because of the advancements in Auto scaling and you know observability and you know, a lot of the predictive, you know analytics and things on the on the operation side. I figured something like that would rank much higher in terms of using machine learning to kind of anticipate.

What's What's coming or what's what's going to happen versus the other key areas, and I know some of these some of these testing functions depending on how you break them out. They can become very Um Niche or specific in terms of skill set. So that might be part of the reason as well.

Yeah. Yeah and just along the lines of you, you know data is data, right you draw the conclusions, you can based upon those that data but there's I think a long history and and you know, I we carry these things around every day that kind of indicates that you know, even when you ask folks what it is that they need and what they think is going to happen. They don't always understand, you know, kind of what they're you know gonna need until they actually see it.

So, you know, we are kind of still dealing in the realm of you know, kind of potential and you know kind of expectation from that, you know standpoint. So, you know, we will kind of caveat, you know the data from from that perspective. And again, I think that both of your perspectives are extremely valuable, you know from the standpoint or saying well that's that's fine date is data and there's you know, 2600 people that are saying that this is my experience.

Right, I would expect it to be different because of this. I think that's you know, significantly valuable as well. Okay.

So Prospect what they're thinking is more like love you doing. So if you think about this like, you know, the automating harder things to automate and that's actually more in the UI, but there are other things use the way I help in this devops or the testing in particular and like the closer to my home ground is okay. So you run all these tests like which testimony like, you know what this or worth while running for this change and they're willing to use algorithms.

Sorry learning in this space and in arguably these things are more useful closer to the left frequencies. I you know that keep on a run day the higher amount of those in ETC. So so maybe maybe these people are thinking about different kinds of way I assisted things.

Well, there was a there was a point in the or question in the chat people were asking. So what it would it. What do you mean when you say AI, you know, that's a pretty broad field like all we deal with that a lot of topics.

I wonder also is oftentimes mom observation is when you're at an inflection point when things are changing or oftentimes when it's still early, there's a lot of confusion right and data also contribute to that confusion of you know, this audience is saying one thing because they're coming out of an industry with a whole different experience than somebody that's living in startup land and building tools for developers or whatever. It might be. You know and maybe we're still in the market hype cycle where it's it's still emerging, right and nobody in snow one is saying like Okay, we don't need testers anymore because it's all AI which ain't never gonna happen by the way, you know versus it's all flipping to something else.

I think it's still we're still in this emerging category. You can argue with me if you feel different. and I believe you're still it's still definitely an emerging category and there's it's starting to only emerge but in all but as well because you're saying different companies implemented and different ways and special and I look at it from the software quality side because there are so many different areas that you can concentrate on you look at just We talk about security at security is a big one you look at you know API if microservices, you know contract testing things of that nature.

So there are so many different areas where you're going to see this explosion of and I look at it from the software development standpoint of which earlier helping the developers write, you know better code. So I think it's there to supplement or augment how we are doing things today. I think we're probably a ways away for it to be you know, you know kind of the Takeover of the machine so to speak but I think what we're seeing though is that there is a lot of Promise in many of these Technologies and you're seeing the areas that you're getting more bang for your buck that we talk about self heal.

We talk about change impact analysis understanding what has been touched on that system. So when I do make that execute those tests and I have to run through Fifty thousand test cases. I can just run through the ones that were affected.

So now I can release faster. So we're starting to see a lot of the the good come out of machine learning as well as AI but there's still a lot of areas where you know, we're going to we're going to see massive Improvement just across the board. Pretty cool.

I think that's maybe a one question. I'm often asked is how do we tell when AI is real and somebody's product and I just scroll down and said, you know ask specifically where are you applying it into what kind of problem and usually you'll get to at least to a domain that will help you and it's not like, you know, we play we just turn AI loose on our data and it comes up with magical conclusions. Probably not right.

We're not there yet. I'm curious to let's talk about where we think so it I think maybe all saying yeah, we're still in this emerging category. And if you're if you're a leader in devops and leader in software development team, maybe you're some maybe you're in management Etc.

What do you think? We should be focusing on as it relates to Ai and our devops environment for the next let's say 24 months. Where should our Focus be when we think about how we apply AI or should we worry about it just renders do it Because God do you mind jumping in?

Yeah, sure think about a lot of things and want to hear what you're thinking about on them. Yes, I guess what I'm going to pitch is like we should be the software engineering should be more quantitative and especially compared to what I see in like a sales and marketing becoming informal data driven. I feel like you kind of puts us in the shame that we are supposed to be the numbers in maths and rational people.

Well, you know, and so and as I think I think there's so that I feel like that's an easier or Peach that you know to get the right help though organizations give the right resources to see you and sister and if not really progress on that data driven that Pro, it's then I think it's only natural that the role the elite that doesn't have to be like, you know, yeah, so that's kind of harm protein it like it's you know, yeah. interesting and yeah, I think you know I would I would totally agree with that and I think for me just as A someone who speaks on this, you know on a regular but just also really a fan even though I work for a better and we're doing a lot of great things with you know machine learning and AI, you know with vision AI within our solution, but I'm seeing a lot of great solutions that are starting to emerge out there even from you know from the from the vendors and I think as organizations want to build their cicd, you know best practices how they can incorporate some things like that mentioned before exam that I'm a huge fan of you know, what KK's doing with launchable right? Because when you just look at what the way they're talking about it, maybe more fan of the marketing side than the other product but but I love what he I love what they're doing because it's really addressing a you know, a problem in terms of this is how AI is going to help devops in emergent right because you're looking at that Predictive Analytics you're looking these different things so One of the things that I see in terms of just the growth of this area is attacking real-world problems.

It's something that doesn't look like it's so far-fetched. I can say we can we already seeing the effects of self-healing we're seeing the effects of change impact analysis, you know, we're saying, you know, a lot of these things, you know come, you know, start to emerge within the software development community. So that's what what I see when I when I look at these the various Solutions in terms of what company can do.

I think it they start with the I think the more tangible things that we can work with with devops that are more realistic which is what we're seeing with some of these with some of the companies that are there on the market today. You know, right? I'm going to share for a third party.

Hold on me. I want it you want to jump in? I'm sorry Mike.

I mean I do I do and and I think part of that is Every organization has to get a sense of their ability to do a lot of custom integration right in early markets. You are always going to have just amount just a significant amount of the weaving together of all these different things falls on the part of the customer. So you have to have the skills.

You have to have the capability you have to have the understanding in order to really make all this stuff work together to impact your environment and a lot of folks go into it and they're just like, oh I can do it and then like, oh, I can't do it right and and then you know, so so being a little bit introspective up front relative about your own capabilities to do that is absolutely critical, right because the reality is the folks that can Will and they're starting to experiment now, they're building it themselves. They're leveraging their own, you know, kind of algorithms. Maybe the vendors help them.

Maybe they don't right, you know wouldn't be the first I'm in a customers come and said hey, I kind of you know Stitch this thing together, it seemed kind of interesting and and The Architects of the vendor go. Oh my God. Oh my God, right and then they, you know kind of started and impact their own roadmap, you know based upon some of that customer based Innovation.

If you fall into that category of no that's not us, right. We need to have this stuff packaged when it's ready. Then you really focus on the outcomes, right?

What outcome do I need do I have a problem in performance testing? Do I have a problem and functional testing do I have a problem in understanding again the data from the op side to to get a sense of what my deploys look like or when I have, you know, operational problems in the stack that will drive you towards which of the solutions make the most sense. If you need that package different environment.

If you don't then you're just building what you need, right but you've got data science to do that. So I don't want to lose that context which is in every early Market. You've got to decide whether you're a DIY type person or you need Kind of the thing to show up at Home Depot and and you know and you just kind of you know, it's it's a kit right, you know, and and I think that's where we're folks need to get a sensor.

Yeah, not in the days of other people did that with the CIA systems? And I know what happened to those products. I'm going by the wayside, you know, if it's a sentence starts out with we built our own fill in the blank, you know, the kind of group you're working with right and folks do that.

Right, you know, this city is the need for intervention. Whatever the thing is, I mean to step on you might there. I wanted to just bring up this one part of the report just to maybe give some insights about where now this question was related to where is AI being applied to but I think these six areas.

It's interesting. They aren't very far apart in terms of rating, you know where they in the report but things like, you know, identifying highest Cricket risk areas automating test cases self-healing Clint. You've talked about that because analysis You know improvements that you can make continuous measurement stuff like that.

So it's not we're not trying to go to the moon right here with with AI. We're it's pretty focused I think on Maybe the software we're building is going to the moon but AI we're focusing on how do we make, you know our software development better and improve our building to deliver better software. So hopefully that's that's a view of pragmatism of how we're applying.

AI to this micro concur with my assessment of that I I do I do and and I think that again it's important, you know for folks especially if they have to do it themselves to be very targeted at what problem am I trying to solve right that's kind of where you know where you're having a hard time automating that's where that question came from. Right? So I think a lot of it is just that ethos of you know, being focused being targeted being discreet in terms of what you're trying to accomplish because as as you said that you know, we we can't boil the ocean at least not up front.

So, you know, you have to be very considerate relative to where you devote those very scarce resources that have the skills to do things like that. are going to wrap things up here and it's been a fascinating conversation and Mike has been fun participating with you and the research you've been in the lead on this and I've had handed in it here and there so it's been great working with you on that what Mitch is really saying is blame Mike. If you don't buy it any of the research all the stuff you agree with is what I worked on all the other well, yes Cascade.

Thank you so much for joining us today, and we wish you all the best. Thank you. And all the fun things, I'm sure you're working on we'll be watching for sure Clint great topic and we appreciate all the world of experience.

tv and you'll see just scroll down a little bit there. You'll see devops on band as a category. You can find it in the series main menu of video programs that we have on Tech strong TV all the episodes that we've recorded.

For text from about devops and bound is on Textron TV. So enjoy that and please download that report. Some great information is free.

The work's been done. It's all packaged up. It's pretty consumable.

I think there's not a lot of reading and a lot of visual content there, too. So we've really appreciate your questions and engagement on chat. I'm not sure if any of us want to take on the quantum mechanics questions KK, you know, you're the smart guy right?

None of us. All right. Well, thank you everybody for joining us today.

com enjoys again, and we shoot the best day and good luck with your AI Pursuits. Good luck with your devops for students. Take care everybody.

Right here. All right. Well Clint kosuke Mitch Mike.

Thank you all so much for joining us today. I'd like to remind our audience of this session was recorded. com slash webinars and just look in the on-demand section.

So at the top of the hour, I promise we'd be giving out for $25 Amazon gift cards. So our winners are Nancy are Lori Z Amelia s And Persona G. So to our four winners.

Congratulations. Please keep an eye on your inbox to clean that gift card. But if you do not see an email just check your spam folder.

So to our audience, thank you all so much for joining us today Mitch has already thanked you and I will thank you once more but we do ask for one extra moment of your time to fill out a brief survey that will pop up on your screen here in just a moment. Um, otherwise, I think that is all we have for you today. We hope to see everyone at an upcoming text wrong learning program everyone.

Have a great rest of your day. provide