Okta's Oktane 2025 | Jenna Cline, Okta

Clips
More from Okta's Oktane 2025

Jenna Cline

SVP of Business Technology

Okta

play_circle_outline Importance of a measured approach to AI amidst hype and concern.

play_circle_outline Integration of technology, data, and AI strategies for better alignment.

play_circle_outline Setting goals to eliminate manual work through AI implementation.

play_circle_outline Enhancing AI Security: Leveraging In-House Products and Trust Layers for Effective Governance in AI Projects

Info
Transcript

Jenna Cline, Okta

Jenna Cline

SVP of Business Technology Okta

Day two coverage of Okta’s Oktane event remains in full swing. Join theCUBE host Rebecca Knight and analyst Jackie McGuire for an exclusive interview with Jenna Cline, senior vice president of technology, data and intelligence of Okta, to unpack how the company is turning AI strategy into enterprise-wide results. Cline explains how Okta launched an AI governance committee before deploying employee productivity tools, setting a measured path for adoption. She outlines the “paved paths” concept that provides teams with safe frameworks to innovate, while sharing... Read more

explore Keep Exploring

What are the ways in which AI is being implemented in practice, particularly in a security-focused company like Okta? add

What is the rationale behind organizing the Technology, Data, and Intelligence teams together in an organization? add

What is the goal of using AI in the organization mentioned in the text? add

What are the different components that build the layer of trust for AI projects? add

bolt Powered by CUBE AI

Jenna Cline, Okta

search

Rebecca Knight

>> Hello everyone and welcome back to theCUBE's live coverage of Okta's Oktane here at Caesar's Forum. I'm your host, Rebecca Knight alongside Jackie McGuire. We've had a lot of great guests today.

Jackie McGuire

>> We have, yeah. I've actually been really, really impressed with the diversity of information. I feel like we've gotten a lot of really practical tips in terms of how to go about tackling this crazy rapid change in the enterprise. So it's been great.

Rebecca Knight

>> Indeed. And here to talk more about that, I'd like to welcome Jenna Cline. She is SVP Technology Data and Intelligence at Okta. Thank you so much for coming on the show.

Jenna Cline

>> Thank you for having me. Happy to be here.

Rebecca Knight

>> So let's start very big picture because I know you've said that what's necessary is a measured and thoughtful approach to AI. But considering all the concern and hype, and opportunity and excitement too around it, how do you translate that approach in practice?

Jenna Cline

>> Yeah, in practice is where it becomes real, right? I think we can all hear the exciting things that AI can do. There's news about it every day. It's almost all I talk about recently, and I think that's awesome. We are living in an age where we are taking another transformational journey in technology, and I'm happy to be a part of it. At the same time, I am very aware of doing AI at Okta at a security company, but any enterprise IT team is going through the same thing. How do we make it real? And Todd talked about it a lot in our keynote this morning, balancing that innovation with the reality of doing these things. When we started out, we took a slower approach. We launched an AI governance committee right away to ask questions, understand, know what we didn't know. And so we did that for a while before we decided to go bigger with some tooling. That took us a few months to really understand what should we be looking for? How should we be thinking about this? Once we had that down, we started with our employee productivity tooling. That was where most of the demand was coming from. A lot of people were using that in their daily lives, and so we were able to solve a lot of the use cases first, just doing that. And then we progressed on and then we learned more. Now we're at a point where we're able to provision a lot of new technologies safely. We call them paved paths, which are tools and guidelines that our teams can use in order to go do their innovation and feel safe about what they're doing. Yeah.

Jackie McGuire

>> That's fantastic. And when you think about, you're the President of Technology, Data, and Intelligence. So one of the things I think is a real opportunity with AI is that as it gives us more productivity, you'll have a lot more data about what your employees are doing, what they're not doing, what they're thinking about doing. So how do you think about that from your perspective in terms of being able to leverage your own journey in AI to really understand how people are using AI and then turn that outward?

Jenna Cline

>> Yeah, that is where it becomes interesting. One of the reasons that my organization now has is called the Technology, Data, and Intelligence is because AI is definitely a technology, and that's how we looked at it at first. But it quickly became obvious that an AI strategy will not be successful without a solid data strategy to go along with it. And so by getting those teams closer to each other, literally within the same org, we could then have conversations about the things that we needed to do to make our data AI ready, and then working closely with my technology and application owners to make sure that that was going to be successful. We also have a team that's looking at AI as a center of excellence. So what can we do to build on top of our technology and our data in order to be able to offer leverageable modules or other things that we can build centrally, but then allow our technology teams to build further themselves? That whole stack and moving that together, the reason for that was to be able to democratize our AI as much as possible. So we're not relying on a single team, we're not having to matrix between organizations on the data side and the technology side. Having that all together is what's allowing us to accelerate.

Jackie McGuire

>> So something that you've said, but also Todd McKinnon said it today on the main stage, that there is immense pressure for CEOs to invest in and deploy AI and get it right too. When you are thinking about where to invest in AI, how do you make sure that you are in fact making investments that are aligned with business goals and business priorities rather than just chasing the shiny object? Because there's also a lot of those too.

Jenna Cline

>> Yes. This is speaking my language. Absolutely. One of the awesome things about AI... Let me back up. When we're looking at AI deployments, in some ways it's not much different than how we look at other technology decisions. It is another technology decision that we have to do. There's two things that make AI a little bit different for me. First is the broad applicability of AI use cases. When we buy other tools, such as ERP or HCM, they're for a very specific area in my tech stack. They might have some other use cases around it, but really we understand why we're buying those things. AI can go anywhere. Should it go anywhere is a separate conversation, but it can go anywhere. The second thing that's different about AI for me in a technology decision is the rate of change. It's changing really quickly. I make technology decisions and purchases within our architecture with typically a three year horizon. Enough time where I know we're not going to be swapping out, but also, we might change because the world will change. AI's changing much more rapidly than that.

Jackie McGuire

>> Every three months.

Jenna Cline

>> Every three months.

Jackie McGuire

>> Yeah.

Jenna Cline

>> So it's hard to say, we're going to stick with this AI. We're certainly not making three-year decisions around it. So these two things are really accelerating and making our decisions a little bit harder, but it also gives us some opportunities. So taking a step back to say, what are we trying to solve with AI? We have a goal in my organization this year to eliminate 250,000 hours of manual work at Okta, and that's on our business process side. So we are working with our finance teams, and our sales operations teams, and our HR teams to identify where there's points of friction in their daily process and something that we can use to either automate, do better integrations, or AI in order to eliminate that friction. And the goal then is to not say, do we need AI? The goal is to say, where can we take out the friction? Oh, and AI is a great use case for that. So it's giving us a business focus on top of the technology decision.

Jackie McGuire

>> So you went about setting that goal, but those hours will be, they won't be eliminated in the sense that you'll need that many fewer employees. We've been clear about this above all the other Okta leadership. But how do you measure that? How do you drive that? How do you hold people accountable? Because that's what I feel like a lot of companies are missing is when they start to roll off these projects, they don't even know what success looks like. So you've defined that in terms of automating manual work. And then how do you drive accountability to that?

Jenna Cline

>> Yeah, yeah. This is not a new problem with AI. Trying to-

Jackie McGuire

>> Accountability is a problem-

Jenna Cline

>> Accountability is one-...

Jackie McGuire

>> for history.

Jenna Cline

>> Yes for history. But the other side is, especially in a software company, what I can see is projects come my way that have a predefined destiny. Like, "Hey, we're going to deploy this tech." But that's not really helping us drive that accountability. The better approach is to deeply understand the problem you're trying to solve.

Jackie McGuire

>> Outcome, right?

Jenna Cline

>> And my organization has, we've reorganized and we've introduced more of a product operating model. So our product managers work with our business and they understand the business well, and they are thinking long term. And their number one job is thinking about what's the business problem we're trying to solve. When we spend more time solving the problem, it's like measure twice, cut once. When we spend more time solving that problem, we design a better solution. And we can be agnostic of that solution that might've been presented to us as something that we should be going after. So we're tackling that organizationally and in our own roles. The other side of it is when we measure that, I do expect accuracy in the measurements. And so we have a number of ways that we're looking at how to achieve that. The first is, I agree, a lot of this productivity that we're seeing, is it something we're actually going to get back? I don't know. A lot of the personal productivity that we're deploying is helping individuals be more productive themselves, which is awesome. I'm not looking at that for our overall goal of where we're removing friction. What I'm looking for is our business processes that we run on a daily basis. We're working with our business teams to understand what does that look like today? We benchmark that. Then we do a solution. We do some planning. What do we think we can save? We take some conservative estimate of that. Then we actually deploy. We do an actual. We figure out where we landed. I triangulate that with my business partners, sometimes with my vendors as well, if they've also been in this. And then I have some confidence in that number that we're achieving is real. And then the more we do these things, the harder thing I think in doing estimates is getting good at the estimate. And the only way you get good at the estimate is to go back and look at your actual and say, did you actually achieve what you were supposed to do? And so it's just a good hygiene for us no matter what we're looking at. But AI in particular, that's really helping us focus.

Rebecca Knight

>> So what's so clear being here at Okta's Oktane is that you are really learning a lot in-house as you experiment and try these different things, and as you said, using these real signals and metrics to know if you are in fact getting the outcome that you want to have. So can you share an example of an AI use case that Okta tried that worked really well, and maybe one where you realized you had to reverse course?

Jenna Cline

>> Where we're seeing it, we are on a bit of a journey. We started with very obvious and simple AI use cases where we would have an AI agent that was acting individually and retrieving information. That was very common. We saw that often, and it was very successful every time we deployed it. What we did when we were building that agent is we built components because it was a very common use case. The only thing that would be different was the context of that agent. And so we were successful in that, not necessarily because the agent itself was successful, more because we built it as composable modules that we could then re-leverage. So if the travel team wanted a bot that somebody could ask questions about it, and also the legal team had needs for that same thing, we built it once and we can leverage it multiple times. And so we went from taking a few sprints to do that to a few hours to be honest. That was a successful thing to us and how we built it. Where it starts to get more complex, that's where we're taking more of a risk to really understand, where we're starting to build agents that maybe are orchestrating other agents or agents that are accessing more sensitive data. We want to be very aware and thoughtful about how we're building those. Our own products are helping us do that. But where we saw some misses was kind of in the in-between where we started saying, "Hey, we feel like this might be a good use case." We didn't spend a lot of time on the business problem. We jumped right into it, and we did the build. The awesomeness about having something that we could then deploy quickly meant we deployed a whole bunch of things more quickly. And looking back and then doing that exercise or going back and saying, "Did we actually achieve it," there wasn't a lot of savings to be had in the first place. And that was another thing going back to let's make sure we're doing the hygiene around it. But all the while, I don't look at anything that we're building as a loss. If it failed and we learned, then we still gained something by it. And I don't want to inhibit my team's experimentation with this because we are building things in a way that I feel good about and the safe and security. So I expect them to go and try new things. And if it fails, as long as we figured that out and we learned what we didn't do right in that one and we applied it in the future, that's a win for me.

Jackie McGuire

>> You've talked about needing to have a trust layer so that as you're building that all of your teams can kind of align and have some faith in that they're going to understand what's being built, how it's running.

Jenna Cline

>> Yeah.

Jackie McGuire

>> So can you talk a little bit about the different components that build that layer of trust that all your AI projects are built on top of?

Jenna Cline

>> Yeah. We start with... This is one of the awesome things about being in an identity company is that we had the products that we could use to go after that. And we were doing that with our product teams too. So when we were deploying things internally, our product teams were learning from how we were doing that, which is really an awesome place to be for an IT team in particular, to be able to have the ability to want to build out this security. And then I have, because I'm an Okta customer too, so I have direct access to my vendor to be able to work with the product teams and say, "Hey, here's working and here's what's not working for us." One of the reasons we started with that simple was exactly that. Can we have the use case where we have somebody that's just retrieving something, and can we make sure that the agent is acting on behalf of the user and only accessing what that user should be accessing? We used our own product, our own Auth0 FGA product in order to be able to do this. We built on that from there, again using our own products, to make sure that these agents are acting and accessing things as we go along. And then also, one of the exciting things that Todd announced this morning is something that I think a lot of IT leaders right now are struggling with is when your AI deployment is successful, that means your teams have the frameworks and guardrails in order to be able to proliferate these tools. But sometimes it's an afterthought to say, "We need to go and audit these tools. Can we monitor these bots?" Like, "Okay, great, we have 1000 bots, but are all they accessing all the things that we need?" And so that's where I think a lot of IT teams are right now realizing, that by deploying these things so quickly, having great success with these, deploying more of these, now we have to make sure that we can govern this new ecosystem that we built. And that's where Okta comes and Auth0 comes into play.

Rebecca Knight

>> Well, you started this conversation by saying that AI is just another technology. I mean, yes, the pace of change is so much faster. But you are, in fact, that's what you're doing is introducing and deploying a new technology to your employees, and a lot of it is change management and winning hearts and minds. So final question, what is your one piece of advice to other leaders who are on this journey, maybe not quite as far along as Okta, about that, about winning the hearts and minds of your employees?

Jenna Cline

>> Yeah, it is just a technology. From an IT practitioner point of view, it's not much different than our others, except for the reason I gave. It's also a big... It's a step change for a lot of teams and doing this as only a technology deployment I don't think is going to set anybody's project up for success. The adoption of these tools, having teams rethink how they do business because of this new capability, helping them with that pace of change and understanding these tools, because they are moving super quickly and I don't expect all of my business partners across the company to have the same understanding that my team might have about this technology. And so we started with the top concerns when you're doing this, what vendors are we using or build versus buy? Making sure we have the governance in place, making sure that we are thinking about the progression of how we want to build these. We did that at first. And then starting to work more in depth on the more complicated use case with our business partners, it became apparent that we also needed an arm that was going to make sure that we did that change part of it, engaging with employees around the company. One of our top company priorities is AI. And so now with that recognition and that communication to employees, as well as the framework that that team is building out, we now also are providing training, specific training for different roles, opportunities to lean more, and then also opportunities to identify things that maybe we haven't seen yet in our own use cases and to close the loop on some bigger opportunities that maybe go across teams. And so having a partner in this so that we can focus on the technology and having a partner that's making sure that we can adopt and drive this, because just because we build it doesn't mean that the users are going to come, right? And so building it and making sure that's happening at the same time was I think the number one lesson to learn.

Rebecca Knight

>> Great advice. Thank you so much, Jenna. Pleasure having you on theCUBE.

Jenna Cline

>> Happy to do it. Thank you so much.

Rebecca Knight

>> I'm Rebecca Knight, for Jackie McGuire. Stay tuned for more of Okta's Oktane from theCUBE. You're watching theCUBE, the leader in enterprise tech news and analysis.