Navigating IT and Business Alignment with Lise Lefaive at OpenText World 2024

This is Textron tv. All right folks. We're back at OpenText World in Las Vegas and we're here with Leash who is one of the leaders of the internal IT operations team at OpenText.

And she's gonna share some insights about things that she's learned along the way. Welcome to show. Great.

Thank you. Nice to have you. Alright.

I guess the First question I want to ask you is, and a lot of people still struggle with this, I think I've been talking about this for three decades now, but are we making any progress in kind of aligning it in the business still? I mean it's still an ongoing issue and uh, today you'll hear folks say, you know, the IT guys, they store the data but they don't know much about the data and the business side creates the data and doesn't really understand where it goes after they create it. And now I got a million copies of everything and costs are through the roof.

How do I kinda bring all this together in some meaningful way? Yeah, it's such an interesting question or topic to discuss. 'cause I remember 10 years ago working in aviation, uh, doing it as well, working on the IT side and saying in 10 years from now there won't be an IT department.

That was my projection. I really thought that there wouldn't, you're still here. I I.

And I really thought the reason why was because data was so proprietary to the business function and unit and use case that in time the business unit itself would run all of its data. And I still subscribe to that thinking to a little bit, but what I've also realized is a cross-functional nature of data. And I think that that's really what makes it challenging, that it is needed.

And they are and they govern and they have to manage and they have to have good data lake strategies. And the worst thing for a company is having an overabundance of data. It's costly, it's ineffective, it makes it hard to um, truthfully find out that source of truth.

And so are we making progress? I believe we are at OpenText. Um, but as I talk to my colleagues in the industry, it's still, it's still an issue.

It's still something we talk a lot about. Um, and so what we tend to do is we really do focus on the use case. The use case, what is the technology trying to solve because data's just an output of that problem.

And when you really put ownership around that use case and you get the people close to that data, speaking to their data, that's the best thing. When you have people managing other use case owners data, they can't really make those strategic decisions with that ballistic thought about what they wanna do with it. And so I really do think it comes down to where is the accountability in the ownership of data in the organization and is it in the right place?

Does this impact the confidence level of the business people too? Because sometimes I'll go talk to them and they're like, they look at the reports they get from it and they go theoretically, yes, uh, I should make a decision that's based on this. But they don't always trust the data that's in their report 'cause they know that maybe it wasn't created with all the data that's relevant at the time.

And so is there a a trust and confidence issue We need to come to A hundred percent because I'm no different. I use dashboarding as a decision maker in my company and I don't blindly believe the numbers or the data that's presented to me. And I will often go deep into just let me know how this happened.

Where do these data sources come, what relevancy when was the last um, update of the information? And again, I find the custodians, the people who use the data the most, who know it when they're super involved with the IT team, proactively challenging being part of the UAT if you will on these dashboarding. And they feel good about those numbers and that data.

That's where the magic happens. It's when you're competing against each other where one group in IT is always providing a service. It doesn't do it for the fun of it.

They're doing it to provide a service and at the core of it, they wanna provide a great service. Sometimes it's hard to do that if the business isn't willing to partner well. And so I really find that it sos partners.

We have some great partners within our own backend, um, who will look at a report and say, this isn't what I, or even better, that's not what I thought the numbers were going to give me. And then you're like, it's a numbers problem, the data isn't right. So, but I love those scenarios because that just goes to show that that person who's using that data already kind of knows the story they wanna tell with it and then they're really challenging deep.

Um, and so I think it continues to evolve, but my experience has been as the people who really believe in their data are the ones who partner best with it. 'cause we too wanna do a great job in giving that right information because On the other side of it, you'll get people who are like trusting their gut, but and when the data conflicts with their gut, they get all bent outta That's right. And yet the reality of the situation is, you know, what they think is their gut telling them something might just be what they had for lunch.

Right? It's true. So how do we kind of increase maybe data literacy among all these folks?

So I always take everything back in it and I've been doing this for 25 years. Go back to what it is the problem you're trying to solve. And you have to work with the people who really understand that.

So I've struggled not just in data, in any type of implementation where the business that I'm providing a service to can't really articulate the problem they're trying to solve with your technology. And that's your use cases. And if you can't have high articulation of that, I really think that it should slow the role with that customer until they can really get really clear with it.

So in in the, in the idea of data or bi or reporting or dashboarding or any of those insights that they desperately want it to, to serve to them, they first need to understand do they have the information that's going to provide them those insights that they can share and we can therefore, you know, give them those insights. Do they know what they're really against? The hardest thing for me when we deal with some of the customers is the cheese moves all the time.

So they're not baselining, they're not baselining their data points. They wanna tell a story, they don't know what they started from. And so I think we in it have to be more vocal as our partner and not just be order takers.

For a long time that's what it was. It was very service oriented. But we need to help our business unit say, you might not be able to tell that story for one more year from now.

'cause the first thing you need to do is baseline where your starting point is and then we measure accordingly. So there's lots of communication but to expect in today's day and age with the abundance of data that the story's gonna unfold on its own. Um, sorry, I don't think it's gonna work.

It almost sounds like we just need more empathy between the business and the IC cell. A hundred percent. And you know, one of those things for me has always been, um, if I'm doing something for a sales organization, I never think about it from an IT perspective.

All of us, I, I'm a sales leader, what am I up against? I have slumping sales. I'm not doing well in this market.

I need every data point to help me. What am I gonna target? What am I going against?

What are the faults? What are the defects? What are the product enhancements?

If I can't rely on that data to tell that story, how am I gonna do my job? I work in customer support. I desperately don't want anyone to call me 'cause I want my products to be so good that no one has to call.

I have to understand that data. So I think that's one of the challenges for us in IT, is take the time to walk in the shoes and be empathetic of what it's like for that person to do their jobs. Perhaps we don't get the same empathy back in.

It doesn't always work that way. 'cause we are seen sometimes as miracle providers. Um, but I really think that makes the difference.

And the minute you as an IT provider can articulate what it what you are gonna do to help the business unit, that's where you really start to see the results. Do you think the rise of ai, may it, you know, and it's scary on some levels, but I feel like maybe AI will force all these issues where we're gonna have to have this moment where we come together and kind of finally have this conversation that's long overdue. Oh My god, this gonna be a better, better, better discussion than what we've been having.

So I'm also running some internal AI and um, I've got lots of business units across OpenText coming in asking for AI to solve. And hardest problem is, is that AI on itself is not magic either. So I I say to the business that comes in with their use case, show me all of your written documentation that we will program these large language models with to help us do this.

But let's just ask one of my business analysts to look at some of this and see if they can answer it. And if the answer is not apparent evident as you comb through some, how do you expect at an LLM to do this? We, we, we still need human intervention.

We still need a human to be able to look at it and say that's the right answer. But what I'm feeling is that there is an enormous amount of potential. We just recently did something, um, with an OpenText with our sales team or for over 10 years they had curated data and the data was told in really nice stories and it was always kept relevant.

And they didn't call blue light blue, they always called blue blue. And they made it somewhat simple and the results out of it easily would shape off 80% of an employee's productivities. I feel like one of the challenges that we're seeing with AI and gen AI specifically is that it's probabilistic and they be overly enthusiastic in its responses.

Um, and yet the processes that we're trying to insert it into our business processes that are deterministic and we want them to be done the same way 100% of the time. Do people understand that? I think there's maybe too much trust in the output in the LLN that people are looking at and going, yeah, all right, that's great.

The machine told me it must be true. Maybe not. Well not only The machine tells you it writes it really well.

Exactly right. It writes it with such authority that you, you might be like, the sky is purple. I didn't know that but I believe it.

'cause it's written in that way. So we have to be careful. I think it can really advance our efficiency.

But without that human intervention layer to say, first of all, that response is applicable to what I was asking. It's not just a really nice response. That sounds good.

We've all seen that when we interview candidates who answer something and you're like, it was really great answer, but it wasn't what I, the question I was asking. So I do think that we have to continuously deal with the human element in this advancement of a technology that is gonna make us much more efficient. Um, but it's not one or the other.

It continues to be a partnership and a collaboration between human and bot. You've been doing this a while now. What are you doing now that you kinda wish when you first started out in this role and doing all this stuff That I don't have all the answers and that no individual has all the answers.

Everything about technology is a team sport. Uh, especially with the info. Like you, you can't do anything in technology whether you're a security person, if you're doing technology right, you have an enterprise architect.

If you're doing that right, you have a solution architect, then you have developers. What about your QA team? There's nothing that one person could possibly know and that it's a team sport.

And I think that as technologists and people who have grown up in tech, you almost feel like you need to have the answers to everything. It's just impossible. It's too vast, it's too big.

So build those networks, hold those networks close to you with them. You can be empowered to do great things, but on your own you can't. I find that, and I marvel at this and how far we've gotten without a whole lot of discipline when it comes to data management.

We're good with structured data, but the rest of this stuff is kinda everywhere. Um, how did we get to this kind of situation that we're in? 'cause it feels a little chaotic and you get businesses function.

So, um, how do we kinda wrap our arms around all this unstructured stuff that's out there that has value? I might argue that people say, uh, data's the new oil, but I'm like, we don't have any processing plants for this stuff. As I structured data is, is a real conundrum for companies, right?

And I mean OpenText has a lot of amazing products to help structure unstructured data as much as possible. But the truth is, is that we have to admit to whether or not this is our future. And I do think I'm a governance person, I've always been in governance.

And so as much as possible I like to normalize data, I like to have it under proper headings. But if the world is about unstructured content and that it's going to continue to manifest and come through, then we need to set ourselves up organizationally to not fight it but instead to accept that it's there set up great departments use the vast technology that can find some sort of semblance in this unstructured data that will allow us, 'cause there's richness in AI and unstructured data too, but it might just not be as simple as the structured data that we look at. But I really do think that fighting the future of unstructured data, I wouldn't go there.

I would more like how do we work around it and how do we create ecosystems? Is all data therefore on its way to becoming hmm, semi-structured. We're wrapping metadata around it and then that's how we turn this.

I hope so unstructured stuff into something meaningful. My brain only works in categories and listen and, and topics and highlights. That's how my brain works.

I've yet, so just in my normal practice and I would challenge anybody, anytime you're given a large set of data, don't you normally try to find, you know, highlights, conclusion, similarities. And I believe that you can, with most data, you can find these similarities. Um, even if that data is presented to you in unstructured form, it just takes more patients longer to comb through it.

But I really truly believe that you can put it all into some form of metadata at the end of the day. Well, Might we get to the point where we maybe, I don't wanna say cast dispersions on people who create unstructured data without categorizing it, but it's getting to the point now where maybe that data's just not gonna be as useful and people will do the right thing automatically. Those people are gonna end up failing.

So let's be frank, right? They're gonna be competing. And I see this happening as round as I deal with use cases.

If someone comes to me and asks for AI to be created in our company and they have strong data control, they understand their data, it's relevant, they know how it's organized and they're compete and someone else comes to me and saying, here's a hodgepodge of 10 years, I don't really know. Where do you think I'm gonna put my time? And so those leaders are gonna find themselves, and I don't mean fail in a bad way, but they might find themselves being irrelevant.

They might find themselves not being high performers because they're not taking the care to get that data right? Um, and so I find that just through time and organically you're gonna see them want to perform like their colleagues are and they're gonna say, what are your secrets to success? And those are some of the key things that'll come through.

Well. So what's your advice to it folks who look at AI and, you know, at least they're told every other day that their job is in jeopardy. But we just discussed this and it seems to me none of this is gonna happen without some IT expertise.

So might AI wind up creating more demand for IT expertise than less? Yeah, I don't see it. I don't see AI replacing it at all.

I see it as just an amazing ability for companies to get to their goals faster. What I do see though is that uh, the people in IT need to be change agents. And if they also fight AI thinking that it's taking their jobs and if they go in with that mindset, it's gonna be a hard journey.

Instead, if we go into, my goodness, this can solve for 70% of somebody's problems and that 70% of people can go do something different that can solve for the next 70%. It's gonna just be an evolution. I think I was hearing this morning, it's the greatest time for change 'cause technology's moving so fast, but there is no doubt that the world of technology is in need of the abundance of people in addition to bots to help us do what we need to do.

So coming full circle, what's your best advice to IT folks about how to make themselves more relevant to the business and kinda insert themselves into that conversation? Be curious. Wanna know the problems that your businesses are trying to solve?

Like truly wanna know, truly put yourself in their shoes. Don't fight the technology. Embrace the technology.

Uh, go in with a mindset of I've never tried to use AI on this use case. What's the worst case that happens? You fail and you go back to your old way, but you've given it a try and then you've learned something from it.

Maybe a year from now you go back and try again because your data structures are better. So I think that it really comes with your attitude and the way in which you embrace it. And if you are going into these situations thinking about 10 years ago, automatically you're gonna find yourself in trouble.

We gotta think about what's 10 years in the future. All right, folks, if you think about it versus or, and the business. Maybe you got the wrong idea.

Maybe it is the business. Hey, yeah. For being on The show.

Thank you so much. All right, We'll be back in a minute folks.