Building an AI-Ready Enterprise - Six Five On The Road

Every enterprise wants to be “AI-ready”—but few have a clear plan for making it real.

‍

From OpenText World 2025, hosts Patrick Moorhead and Daniel Newman are joined by OpenText’s Shannon Bell, EVP, Chief Data Officer & Chief Information Officer, to unpack what it actually takes. They cite new CIO research while exploring the role of unified, governed data and how deeper collaboration across the business—particularly between HR and technology—becomes the differentiator as AI embeds itself into every workflow.

‍

Key Takeaways Include:

‍

🔹Bridging the ROI Gap in AI Investments: Insights from recent CIO survey data reveal what separates organizations seeing AI-driven returns from those still navigating the early stages of deployment.

‍

🔹Foundations of AI Readiness: Proven methods for improving information quality and enforcing data governance–critical in supporting scalable AI adoption.

‍

🔹Value of Unified Data Platforms: How integrating analytics and information management delivers measurable operational gains for enterprise IT leaders moving to platforms like OpenText’s Data Cloud.

‍

🔹Human Expertise vs. Digital Agents: Key strategies for effectively combining human insight with the expanding use of digital agents in core business processes.

‍

🔹Cross-Functional Collaboration: The evolving partnership between HR and technology in managing AI agents—covering job design, onboarding, measurement, and team integration.

‍

Learn more at OpenText.

‍

Watch the full video at sixfivemedia.com, and be sure to subscribe to our YouTube channel, so you never miss an episode.

‍

Or listen to the audio here:

‍

‍Disclaimer: Six Five On The Road is for information and entertainment purposes only. Over the course of this webcast, we may talk about companies that are publicly traded, and we may even reference that fact and their equity share price, but please do not take anything that we say as a recommendation about what you should do with your investment dollars. We are not investment advisors, and we ask that you do not treat us as such.

Patrick Moorhead:

The Six Five is On The Road here in Nashville for Open Text World 2025. Daniel, the big topics here have been two of our pretty much favorite topics, agentic AI and data. And here we were in the audience and Open Text brings out a data platform.

‍

Daniel Newman:

I thought you were going to say like, oh, agentic AI, AI, generative AI, AI on data, machine learning. No, I'm kidding. But yes, it does feel like it's been a whirlwind of a few years. And it's great to come to these events where enterprise AI is really put into focus. You know, you and I, I think it both really said the next growth frontier for AI is going to be bringing it into the enterprise. Unearthing the value that sits inside of the proprietary data. The number thrown out here was 90% of data sitting behind the firewall. I've heard numbers as high as 95. I've heard other CEOs say that 99% of data has not touched AI yet. Right. Enterprise AI. enterprise data.

‍

Patrick Moorhead:

So it's exciting. For sure. And, you know, when I talk with CIOs, same with you, and we do our body of research, what keeps coming back as an impediment to do agentic AI, it's getting your data ready. And whether that's making sure it's clean to give you accurate results, making sure governance is there, and about 10 other things just keeps coming up. In fact, I did a CIO roundtable in London two weeks ago where, yeah, that was their number two issue. And I can't imagine a better person to talk through this than Shannon. Welcome to the show.

‍

Shannon Bell:

Thank you.

‍

Patrick Moorhead:

I know you have two roles here, and I'm sure you get two paychecks for CIO and Chief Data Officer. It's a twofer.

‍

Shannon Bell:

It's a twofer. I am responsible for supporting our 22,000 employees and our 330,000 enterprise customers. So a lot on my plate, but I have an amazing team to deliver.

‍

Daniel Newman:

Wonderful. She handled that very well. No, didn't she? She was so ready for that question. Yeah, well, she didn't make the normal joke about them, I feel like I deserve two chairs. That's what we normally do. My boss would never. I never take care of my boss. Your boss is terrible. I never take care of him. So anyways, you did some really interesting research recently, CIOs, in this survey. You were looking at what I think everyone is right now, AI ROI, right? And kind of curious, especially in enterprises now, as we really have seen the tables turn, the market, the whole sentiment is, okay, AI is cool. Time to start telling us where the value is in this thing. What did your research show in terms of organizations that are able to determine real ROI and others? What are the big differences?

‍

Shannon Bell:

I think the big difference is a lot of people first approached AI as a technology-driven challenge. And so it was really implementing technology for technology's sake. And AI is pretty amazing technology. But if you don't apply it to a business problem, you're not going to get the results. And that's really been the learning and what we saw through the research. Where AI was applied to solve a specific business problem, where the value and the quality of the data was known, where the business processes were well established, you could see material outcomes and benefits. Where you were applying AI because you could, because it was a cool technology, you weren't seeing the same value and outcomes, and that's really what we're seeing across our customer base.

‍

Patrick Moorhead:

Yeah, it's interesting, Gannon. You did an early body of research that talked about the success of programs and how the CEOs were involved, and then management as well.

‍

Daniel Newman:

Yeah, I mean, we knew coming into this year that this was this big moment for AI. You're talking about the research we announced at Davos last year for CEOs. And I think, you know, when you get into the CIOs, the CEOs and the boards really are now, you know, this is being delegated to the offices of the CTOs, CIOs, CDOs of organizations and saying, OK, I've got a mandate. You know, whether you're a public company, private, the shareholders expect us to have an AI store. Now we need to basically say, okay, whether it's a CapEx investment, whether it's OpEx investments, whether it's changing in your hiring strategies, whether it's the innovation in your products, the expectation is that people want to not only know that you're getting behind AI, because that was kind of table stakes. Like two years ago, it was like, you added it to your earnings remarks. You said AI, and that was good enough. Now it's like, what's it doing? So it sounds like really as we move from CEO to CIO, that's really where that answer is going to be on earth.

‍

Patrick Moorhead:

Yeah, so one very consistent theme that we've seen is data, and not just data, but information to get good outcomes from agents, or even if you're doing a simple front-end LLM, And it makes sense, right? Garbage in, garbage out has been a thing since the first IBM mainframe. It really has evolved. And the way that, or at least what I'm hearing is that, hey, it's one thing to get, let's say your ERP data clean, okay? But to connect the front end to the back end, all the M's in the enterprise, that's a vastly different type of exercise. And you can't just throw it in a data lake and call it a day. That has actually never worked. I've seen people looking at data fabric architectures, things like that. What are some of the best practices or approaches that you've seen with your customers and even internal to OpenText for the best outcomes?

‍

Shannon Bell:

We're in a fortunate position as a leader in secure information management. It means that our customers are already seeing the value in investing in information management. So layering AI on top actually delivers really good outcomes for them. That being said, if you don't have the systems and processes in place. And that's kind of the key. It's one thing to have the data and know the data. It's important to have the connectors into all the systems to pull that data. But having good business processes means that you can actually leverage the data and drive agentic capabilities and outcomes. And that's where we've been focused. And we treat ourselves as customer zero for our own technology. And so it puts me in the unique position of being the first customer to deploy our solutions. And what we've learned through 70 plus implementations of our different products is that where we have focused business cases, where we understand the quality of the data and have good governance around the data, we can drive very rapid outcomes. And I'll give you an example in our service management space. Last year, we invested in consolidating our system. for all of our internal help desks. And we saw immediate benefits just through consolidation of our systems, 30% benefit in terms of reduction in L1 tickets. A year on, because we made that investment in our processes, we were able to layer AI on and various use cases and saw a 70% reduction in our L1 help desk. And so the investment in good information management, governance around it, knowing your data, layering in the process pieces, and then tackling AI from a position of strength gives you those outcomes.

‍

Patrick Moorhead:

So one of the keys here seems to be the connector. And just, you know, to edify everybody, including me, are we talking APIs, MCP, ETL, things like this, copyless type of scenarios?

‍

Shannon Bell:

We're talking about all of the above. So we're talking about API to API integration. Absolutely. We're talking about agent to agent integration. We're talking about the underlying backend data stores and how you're bringing those data sets together. So it's integration at multiple layers. If you think about the evolution of enterprise content management, it's really been an evolution of bringing together different data sources in a governed way to meet the needs of the enterprise. Now that same investment can meet the needs of enterprise AI.

‍

Patrick Moorhead:

I appreciate that.

‍

Daniel Newman:

So you gave a couple examples and I really like it about kind of where you're seeing gains. You know, you mentioned governance, you mentioned information management, analytics too. And again, you're not doing it all. I think, uh, you know, we talked to your president of worldwide sales, Todd, he was talking a little bit about ecosystem collaboration. I mean, that's a really big part of this, but like as you put all this together and they're adopting the model of sort of not necessarily you know, one going maybe all in one solution, solves everything, partnerships, right? They're going to bring governance, data management and analytics together, then layering AI on top of it. Like, where are you seeing the enterprise leaders that 300,000 that you said you support these customers, like, where are they getting the biggest gains?

‍

Shannon Bell:

So they live in a complex ecosystem, first and foremost, and we live in a complex ecosystem in our own organization with many different applications and systems. And a part of the challenge is that, you know, there's many use cases you can tackle and it's how do you prioritize to get the biggest gains? And when you think about implementing agentic AI, I always say, start super simple to the point that you think it might be too simple because where some of these enterprise AI pilots and projects have failed is they went for the most complex use case first, which had a lot of corner cases, was prone to fall out. and didn't see the value. And so we start very simple. We start with customer support, with help desk use cases, with developer efficiency use cases, use cases that you know there's going to be value that you can measure, and use cases where there's a strong baseline, highly repetitive, lower value tasks, high volumes in those tasks, things that are easy to actually build into discrete agents. And then you can start to build the complexity and the orchestration of those agents. And so we do it in our operation center. We have agents that are performing discrete tasks to get to the root cause across problems and incidents. And we have an orchestration layer that does a little bit more complex use cases where we're willing to have some fallout. And so it's really a matter of picking those use cases where you know there's a high benefit to implementing a level of automation and AI. And that's what our customers are doing as well. They're bringing us those use cases and saying, start here, let's prove the value, let's build the confidence internally around this agentic workforce, and then we can grow into more complex use cases.

‍

Patrick Moorhead:

Yeah, we're seeing industry-wide a lot of success with generative AI and agents across coding, across customer support. Anytime that there's a document source that can be pulled in, and even very simply queered against. That could be for human resources as an example. And also a lot of success in marketing. We talked to your chief marketing officer and she talked about a lot of the collateral that she was doing and even brand campaigns that she was leveraging AI for. So you did a really good job talking about how you break down the complex complex things, simple things, but how are you balancing kind of the human in the loop here? There is still a human in a lot of these interactions.

‍

Shannon Bell:

Which I think is the most critical piece, because the agentic workflow should not be replacing humans from making some of those decisions. And so we've been very thoughtful about how we bring those pieces together. We actually write job descriptions for our agents. And so alongside our human job descriptions, we actually have agentic job descriptions that have the input, so what data sets are being used, the outcomes, the expected outcomes, how it's being measured. And if you think about the life cycle of onboarding a human, we look at the same life cycle for onboarding our agents. The job description, the outcomes, the governance, the measurement, the offboarding is equally as important. I would say we don't go as far as putting them on a performance improvement plan, but when the performance is not there, we're not shy about shutting down those agents. And so I think managing them very thoughtfully and understanding the role of the human in that life cycle is critical.

‍

Daniel Newman:

And just elaborate on that a little bit, like when it comes to job design, when it comes to, you know, the augmentation, there's a lot of sensitivity right now. It's a big topic. Obviously, you can get a ton of scale with agents, but performance, you kind of indicated like, again, you're not doing performance reviews on agents, but there's gotta be like a certain threshold where they're showing that creating agents and building teams around and with agents adds more productivity and efficiency gains than say, doing it all with. Like, how do you sort of balance that? Like, how do you balance the HR part of that? How do you balance the data and the measurement part?

‍

Patrick Moorhead:

By the way, who do they even report into? I've had a conversation with a lot of CIOs that are like, well, wait a second, where should the marketing agent report into or the HR agent, right? And it's not a political thing. It's more of who's in charge of this agent and what they got done and the quality and the speed at which they did it. And a fun bonus question is, do agents ever report to other agents?

‍

Shannon Bell:

They do. You have orchestration agents. So it's complex. But I think the role of HR is also evolving. And I would say we haven't gone so far as to call our human resources team our human and digital resource team. But you can see where it's evolving. And we worked very closely with our HR team as one of the first adopters of the technology. because we knew that they would be instrumental in helping drive the change management across the company. So they saw the value, they understood, and they worked with us to drive the change management, which I think is really critical because having people understand what this agent is expected to do, how it impacts their role, and where the opportunities are, roles are changing and evolving. And so if you think about an agent performing routine tasks, perhaps the person that was doing those routine tasks is training the agent, is training the next set of agents, is building out those capabilities. That requires re-skilling and new knowledge acquisition that doesn't exist widespread across the industry. And so a lot of investment in training and having people take the fear away from what the agents are doing by being very clear about what the job is. That's why we've invested in the job descriptions. It helps us with change management. And it helps us with people owning and adopting the technology.

‍

Daniel Newman:

For sure. Well, it certainly simplifies benefits, you know. Well, just make sure they get enough GPU performance. Now, if they become sentient and they start asking for, like, 401Ks in healthcare, then we've got a scale problem. Shannon, thank you so much for taking the time here with us at Open Text World 2025. Hope you have a great event and look forward to checking in again soon.

‍

Shannon Bell:

Thank you very much. Thanks.

‍

Daniel Newman:

And thank you everybody for being part of this episode of The Six Five. We're on the road here at OpenTexas Rural 2025 in Nashville, Tennessee. Subscribe, be part of all of our content here at the event. And of course, all of the great content across The Six Five. But for this episode, it's time to say goodbye. We'll see you all later.

‍