AI-Powered DevSecOps for the Enterprise – Derek Holt, Digital.ai
What is it? AI-powered DevSecOps for the enterprise is not just another innovation; it is the catalyst for scaling and accelerating innovation across the enterprise. It is the fusion of three AI capabilities that until now have been implemented ad-hoc in the development and delivery lifecycle: AI-powered software delivery workflows, AI based code governance, and Predictive Intelligence. How it is unique? AI-Powered DevSecOps for the enterprise is the first holistic approach to creating efficient, automated, intelligent, continuous software delivery processes for developers and teams of teams, across the enterprise. Leveraging leading-edge AI-based capabilities and implementing them at enterprise scale, it exponentially
increases the opportunities for transforming DevSecOps into new sources of business value and competitive advantages.
Transcript
This is Techstrong tv. Hey everyone, it's Alan Shimel back here on Techron tv. I'm really happy to, uh, have a, a guest on today who I've wanted to have on for some time.
You know, we've been covering digital ai, I think from the very day it was launched. I, I remember, um, and you know, Derek is going to give you a little of that background, but Derek Holt is the c e o of digital ai. It's, it's great to have him on here and can't wait to, uh, talk to him about it.
Derek, welcome to Text Trunk tv. Thanks, Alan. Uh, super excited to be here and looking forward to, uh, catching up.
Absolutely. So, you know what, before we jump into the digital AI story, Derek, we, we really should talk a little bit about you, right? I, how long have you been c e o over at Digital AI and kind of, you know, what'd you do before that?
Yeah, so why don't, why don't you share a bit? Absolutely. So I, I've been with Digital AI since, uh, summer of 2020, so really those early days that you highlighted as we were forming the company, and we can talk a little bit more about sort of that, that background.
Um, I came in originally as a general manager for a kind of a subset of the, of the company and then, um, uh, became c e o, um, here at the beginning of, of, of this year. So, uh, it's been a, it's been a super exciting run. I'm, I'm a software guy by training.
They don't let me write code anymore, which is probably good for our customers, but, uh, I still do it on the, on the, on the weekends. But, uh, I got my start, uh, early days for those that, uh, remember the early days of Eclipse. I was part of the, the, the, the team at I B M that was working on, that worked with, with the rational software, so the early days of, of, of Agile transformations and, and into early DevOps.
And then, um, spent about a decade, um, in, in kind of 20, uh, 11 to 2020 doing, uh, different venture backed starters, but always in software, always in enterprise, and always been, uh, just a, a huge fan of, of the space that we get to both live in, uh, every day. Yeah, Absolutely, man. And then, I mean, you, we, we've kind of beat around the bush a little bit about digital ai, but digital AI was originally formed as the result of a combination of a few, you know, pretty like best in class Yeah.
Uh, companies within the DevOps space, right? Yep. Yeah.
You, you, you know the story better than, I don't wanna put words in your mouth. Why don't you kinda, yeah. It's interesting.
So if, if you, if you go back to 2019, I think the opportunity that we saw in the marketplace was, um, uh, to build a purpose-built company that focused on enterprise customers, uh, in this DevSecOps, uh, space, of course. And that, that meant we gotta do these things at scale, right? We can't just be agile at the team lever, or have one pipeline that is automated.
We gotta be able to do this at, uh, very, very complex, uh, globally distributed scale. Second, it meant, uh, heterogeneous environments, not just heterogeneous tool chains, but also we need to be able to build capabilities that support the realities of large scale banks and insurance companies. And that's 40, 50, maybe 60 years of software archeology, right?
You got everything from the mainframe all the way through the, the latest in cloud and mobile and containers in Kubernetes. And then lastly, we saw the opportunity to also, uh, be a little bit more, uh, forward leaning around data. And, and, and you and I were talking about this a little bit.
Obviously, you, you, you can't, uh, turn on a, a, a TV show without talking about ai, uh, these days. But we had digital AI in, in our name, well before the latest wave in large language models, generative. And the huge part of that thesis was when you do these at, uh, these, these business processes at scale, right?
You have a tremendous amount of data, but historically it's been locked in these product specific domain specific data warehouses. And yes, I can give you a burn down chart, and yes, I can give you some views into how our testing is progressing, but nobody was able to give an end-to-end view both how you build software, are you good at the process of building software end-to-end? But also did it matter?
Was the value being delivered to the customer? So bridging not only all the tools on the development side, but also, uh, in, in production, and frankly, assuming, which I think is really, really critical, kind of focusing on the enterprise, we assume heterogeneous environments because that's what we run into 100% of the time. And so it allowed us to kind of, uh, uh, take that view and say, Hey, we're not only gonna support data sources that are our own, but hundreds of data sources and, and, you know, integrations beyond that, which again, is the reality of, of, of the enterprise Sure is sure is.
You know, I I, I'm listening to you, I'm shaking my head saying is all true. com, it's coming on 10 years, and we've, we've seen this, uh, evolution transition from DevOps to DevSecOps, right? I, truth be told, I got into DevOps because I'm a security guy.
I've been in security for 25 years. I thought it was gonna be great for security, that was why I did it. But now we've really seen it kind of full bloom, right?
DevOps is DevSecOps, and it's the same thing with digital ai, right? You guys are, are all about DevSecOps and, and how important it is not just to the security people, not just to the DevOps engineers, not just to the developers, but to the entire organization here. Now, bringing this full circle to AI and, and what we've seen, you're right, you guys are always had AI in your name from day one.
And I'll, I'll testify to that. And that, that was before people started. Were talking about generative AI and ll uh, you know, large language modules and so forth.
But there was a lot of, you know, back then, I call it kind of the machine learning phase of ai, if you will, the ml AI kind of stuff. Yep. And just because we're all, as an industry, we're very shiny trinket focused, right?
Just 'cause we got this new shiny trinket doesn't mean that ML AI kind of, uh, technology was bad or gets thrown out with the bathwater. Right? I wanted, I want your opinion, Derek, is, is when we look at DevSecOps and AI now, right?
'cause everybody's talking about that. Yeah. Talk to me, what, like, what AI are we talking about?
Some of it, all of it, which parts of it, what, where, what's real, what's not? Yeah, so I, I think there's a couple pieces there. First of all, I totally agree with you.
ai and, and a lot of other companies, we have been saying Dev SecOps for a long time. It's been very clear to me over the last 18 months or so, that has now become the, the, the, the label everywhere, right? And, and, and in some ways it's just one of those things as things continue to transform.
But it as always, it's about automating and, and having repetitive measurable processes to make it easier for us to deliver high quality, high secure, et cetera code, uh, to market. So those, those market pressures have only increased and say, we gotta continue to do more and more and more, but we gotta be able to go fast with confidence. And that's one of the other pieces for us, like, because, and I've always thought about this as a real, um, almost a superpower, is that we get to focus on a subset of the market.
That's not to say like, we love startups, we love small businesses, but we wake up every day thinking about the enterprise. And when you have the luxury of just thinking about that one set of, of, of, um, of, of entities, it allows you to make some decisions a little bit more quickly and a little bit more obviously, around what does set dev SecOps mean. Not just generally, but what does it mean to a large scale bank or a large scale insurance company, or a, a government entity.
So I totally agree with you that we've kind of flipped the page there and, and, uh, I'm excited. 'cause I think you're gonna also see, uh, uh, more and more quality automation, uh, you know, coming in, especially with, with this, uh, pending acceleration of productivity that may be coming from the, um, large language models and generative. But to, to answer your core question on ai, and I think about this very, very simply, um, what is the right data mechanism to solve a real business problem?
So, to me, whether we're using regress a simple regression, or we're using machine learning, or we're going as far as using large language models and, and generative, it really does have to go back to what's the problem we're trying to solve, and what is the right instrument to solve that problem? I was just on, uh, this week with a, with an executive at a large scale customer of ours. And, and they were describing that like, in some ways the comment was like, who cares if it's AI in the context of, of the latest generative?
Does it solve the business problem at hand? Now, great news is we got a whole bunch of additional tools now at our disposal, and all of these different elements have a role to play. But, but clearly your your point is, is well taken, which is, some of it is a lot, you know, there's a fair bit of hype.
There always is. I think we all can agree that, um, the notion that we're gonna see pretty significant productivity gains that we're gonna see pretty, pretty interesting evolutions around, uh, the topic of citizen development that we've talked about a long time, or we talk a lot about citizen, uh, uh, testers, right? Where we can just use natural language to generate test, uh, you know, automation test cases.
I think we all know that's gonna happen. I'm sure there'll be the hype and then the fall, and then the, and then the stabilization that we, that we've always seen. But for us, it's about how do we make nachos individuals more productive?
But really, really importantly, and I think this is, uh, critical, how do we make teams and organizations more productive? Because a lot of the announcements, a lot of the hype to date has been really around how do I make a individual developer, an individual tester, et cetera, more productive, super important. But when you think about that, how do I then get that to production?
We've got a lot of customers that today that are struggling to take the human generated code all the way to production in a timely, secure manner. Imagine if we've got 3, 4, 5, 10 x the amount of code that we've gotta get into production, it puts a tremendous amount of strain on those pipelines. So getting folks ready for this pending, uh, kind of industrial revolution type productivity enhancement, if you believe what, what what we're reading these days, I think is really, really critical.
Absolutely. It really is. Hey, Derek, I don't want to make it, you know, this a product pitch, but let, let's do some real life examples of how, you know, all these different types of AI and so forth are finding their way into, let's say, the products at digital ai.
It's a great question. So we think about it, and we actually just made an announcement on this, uh, or earlier this week. I think the, the days are all blending together, but, uh, certainly, uh, over the last few, few days, um, we think about it right now in three big, big buckets, right?
So first and foremost, um, we're thinking about, um, how do we use these AI solutions to kind of power, power workflows that everyone's doing? So whether you're in our enterprise agile planning tools, or you're in our testing tools, or you're in our, our release orchestration deployment automation solutions, our, our security, our application protection tools, how do we use AI to accelerate? So a great example there is one of the things we, we, we, we highlighted this, uh, this recent announcement was being able to take freeform text and automatically generate automated test scripts with it, right?
So just, we call it text to test very basic test creation. Um, and, and that, that notion is gonna accelerate those workloads, right? Because as we all know, um, uh, we often run out of time testing more than we ran outta all the tests.
Uh, when we think about the way that quality is thought of, and we're doing that in the planning phases, we're doing this in the threat analytics phases, we're doing it in things as simple as our knowledge assist. So instead of searching our knowledge base, you'll be able to talk to our knowledge base, right? For lack of better terminology.
So I, and I think, look, I think all of those are very, very important. Um, the teams are demoing some really, really cool stuff that's coming after it. Um, but I think everybody's gonna do those things, right?
Yeah. Like, that's gonna be the reality. You're seeing it across the board.
ai, the first is governing AI based code, uh, that was generated, right? And, and the idea there is all of the best practices with, with DevSecOps that that, that you and I have been around for, for a while to watch. Um, they were best practices when, when humans were developing the code, they are now must-haves when, when we have code that may have been generated, uh, beyond and, and it's now not just security and quality, it's compliance, it's governance, it's other elements to make sure that we can take advantage.
And we talk about internally, how do we responsibly take advantage, take take advantage of the benefits that are coming from ai. So we highlighted a whole set of capabilities in our, in our core dev SecOps portfolio that are sort of tuned into understanding what it means to bring generated code or at least a subset of the code, uh, that has been generated, you know, to production safely, responsibly, uh, et cetera. And then lastly, really the, the, the, something we've been doing, you know, going back, you mentioned, we formed a company through a series of acquisitions, a company called Numerify that, that has been doing this for, for the better part of the decade.
We, we announced a series of enhancements to our predictive intelligence solutions. So this is breaking down those data silos as you highlighted, and then being able to give not just 360 degree rear view views across tool stacks, across production and development across different domains, but also to predictive future based on the past. And we've been doing this at scale with large scale, uh, uh, uh, you know, fortune 100 type companies for, for years now.
But it's around our change risk prediction model, our service management process optimization models, our, our flow acceleration model, which is one is getting a lot of, a lot of excitement now, but it's about using that data to make better business decisions as we, as we drive things forward. So, you know, it's one of the things we talk about internally. We want to paint a vision of where this is all going, but we're doing this stuff today.
I mean, we're working with, uh, maybe since we last chatted, but 50% of the Fortune 100, over 50% are customers of ours, right? Eight out of the 10 largest banks in the us, largest, largest eight out of the 10 largest banks in Europe. It goes on and on and on.
Um, and we're just super excited to be able to work on these problems at scale and, and in these complex environments, right? Yep. You know, I'm listening to you though.
Part of me shakes my head because yeah, you're like, yeah, no, everybody's going to be doing this. Everyone's doing it. Sometimes we forget, Derek, we live in a bubble, man.
That's true. You live in a bubble. That's true.
I live in that bubble too, where of course, we're using these AI tools to, to talk, to talk to our knowledge base and to, right, and to spin up, you know, uh, test based upon just natural language, you know, stuff. You know what, Derek, I'm gonna venture to say that maybe 15%, maybe 15% of vendors are doing that right now, and maybe another, I don't know, 35, this class of crossing the chasm, right? Yeah.
Yeah. We're an early adopter land here. That's right.
Let alone that, that first, you know, the first, uh, 30%, which is that first early mainstream, if you want to call it. But we, you know, this is still radical. This is still kind of brand new for a lot of people in our audience.
And it, and it's crazy. I mean, the fact that you can talk and, and basically spin up testing based upon what you tell it you want. Yep.
Yep. It's pretty damn cool. Pretty Damn cool.
It's cool. I, I, I, I think too, the, the part that I, we do have to remind ourself, 'cause you're right, we, we always have the latest ga this is our career, right? This is what we wake every day thinking about.
It's our world. The biggest, the biggest thing that I, I am eci it makes me excited about it, is I have this absolute pleasure privilege to spend time with a lot of CIOs and CTOs as I know you do, right? Mm-hmm.
And, and we get to, we get to kind of see what's happening in reality, right? At a large bank or at a large insurance company where it's much more complicated, right? And I think this is one of the unique things.
I've done a series of venture-backed startups in my, and the, the decisions you do when you're a venture-backed startup, and there's 10 of us, we can go really, really quickly, and if we make a mistake, it doesn't really matter, right? And there's a, there's a lot of that, that piece. But doing these things at scale is hard.
And at the same time, I have been really, really excited to see some of the momentum that that is building in the market around companies that are a hundred, 150 years old. If you, you know, if you go back in, back in time, and one of the things I always kind of talk to the, the, the leaders there about it, and I think this holds true, is this isn't a binary, right? This is a continuum, right?
And, and it's a journey, not a destination. And so part of the advice I would give if anybody was so inclined to ask me for advice is get started, get started on a small piece of it, because you're gonna get benefit and it's gonna, it's gonna stack on top. And this wave is no different than when we had the early DevOps waves.
No different than we had the early agile development wave. You've just gotta not get kind of, um, stuck and not get started. You just gotta take one foot in front of the other and benefits come.
I I, I'll be honest, like you, we've, we've been through a lot, right? Yes. I remember the DevOps wave, obviously.
Look, I remember when everyone was talking about cloud and you went on a VC pitch and they said, well, what's your cloud story? Right? Yeah.
That story. Exactly. Yeah.
Yeah, exactly. Right. But this feels different.
This feels more like when the internet first started coming out. That's right. Where it's not just gonna change people like you and me, Derek, this is going to change.
This is going to change all your relatives who call you up for IT advice all the time, right? Yeah. They, they're gonna start talking to their thing.
And I'm not talking about sir stuff. I'm talking about real, you know? Yeah.
Talk to my phone or whatever my device is and, and get stuff done. And it's gonna have a civilization wide impact, not just our industry wide impact. Yeah.
But companies like digital AI are gonna be out in the forefront on this, on, on us utilizing this to make your core mission, which is making your users more secure, faster, better. That's right. It's gonna make them, it's gonna bring it to them faster.
And I think that that's the takeaway here. I, I think that is a big, big takeaway. Now, I, I, I, I'm okay if I get less tech support phone calls from my family members.
So if we can figure that out, that's a great, great situation. Amen. Amen.
But, but, but look, I, I think we wake up every day, every day thinking about our goal, which is to help enterprises unlock these digital possibilities, right? As they, as they think about it. And we spend most of our time with, um, folks that were around before the internet.
So, you know, they, they, they, they're more, uh, traditional, uh, enterprises, but, but have woken up now and, and, or really over the, over the last few decades and realize, holy cow, I'm a software company, really, and I just happen to be in a, in a certain industry. And, and it's so, uh, to me inspiring to see them thriving and, and, and, and driving the market, the market forward. But no doubt as with, with all of these waves, there's gonna be some, probably some, you know, job destruction.
There'll be some job creation that comes from it as well. Um, and, um, and, you know, we gotta, we as a, as a broader economy, gotta kind of make sure everybody's paying attention to that. But, you know, for me, um, everybody should at least be aware of what's going on, should see how they can use it in their day-to-day workflows.
Because to your point, um, you know, there was a point in time when folks said, Hey, this internet thing was a fad, right? And that is not been the case. Yeah, yeah, for sure.
I was talking to a VC the other day, we did an interview and he said something that really hit this home for me, and I think it's the perfect cherry here, which is, you know, 10 years ago we were saying every company's a software company, right? And software's gonna eat the world today. Every company's an AI company.
Yeah, yeah, yeah. Right? And digital AI was out in front, right?
You had the name four years ago. So, but that's, that's where we are. ai to get started.
Absolutely, Absolutely. More. Absolutely.
Yep. You can, you can go to our website. Um, there's a bunch of great AI content there, there's a bunch of, of the, the new announcements that we made.
And again, al you know, huge thanks for, for having us on, for, for what you, you all do to continue to push the, the industry forward. Uh, it's an exciting time. I'm sure there'll be ups and downs and hypes and, and, and kind of the trough of disillusionment, I think is the terminology that is often used, but yeah, it it's Not always the hockey Stick, right?
Exactly. But, but this is, I I think I'm with you. This is, this is a real inflection point in the industry and it's, it's a, a privilege to be, to kinda have a front seat at it, right?
Yeah, no, well, look, you guys are leading it. You were way out ahead on this. Good, good for you, man.
Keep it up. Come back, come back and keep us posted on this. Okay, Derek, You got it.
You got it. Have a good one, Alan, thanks so much. Thank You.
Derek Holt, c e o digital AI here on Techstrong tv. com, security Boulevard. We're gonna take a break on Tech Drunk tv.
We'll be back in a minute.