CTO Chats with Fernando Cardenas, CTO at MBO Partners

Fernando, please tell us about yourself and your role.

I work at MBO Partners as the Chief Technology Officer. In that role, I’m responsible for ensuring our technology is generally working and that we have a roadmap for upcoming IT developments.

Given that the IT landscape evolves rapidly, what challenges are you currently facing and how are you addressing them?

Today, the pace of innovation is so rapid. The question is, how do we best utilise all the different options to our benefit? Given that our world has changed so quickly from a business perspective, how do we help the business improve? Our business landscape is also changing quickly. It used to be that the technology landscape was changing fast, but the business landscape was not quite as fast. Now they’re both changing quickly. How do we, as a company, work together to understand the business landscape changes and create a strategy that aligns with that, given that the technology landscape is also changing quickly? It used to be more of a focus on the technology side, but now it’s a partnership between the business and the technology side.

I’m also curious about your vision for the future of tech. Do you think the IT sector will evolve in the next three to five years? What major shifts are on the horizon?

I would say a couple of very large shifts. The biggest one is AI. If you’re not paying attention to AI, if you’re not doing something with AI right now and understanding how it will work, it’s like not doing anything on the internet in 2010; you’re insane not to start using it. That’s the biggest shift.

Then the second is, now that you’re using AI, the question is, what’s the responsible way to use AI? Those are the two biggest things. First is figuring out what we should be doing in AI. Then the second is, given that we should be doing something in AI, practically speaking, how are we going to roll that out across the organisation? Are we going to do pilots? Who benefits? Because it’s not a technology, from my experience, where everything works great with everybody. It’s very much tool-focused right now. You’ve got a tool for recording meetings, a tool for selling, a tool for marketing, a tool like Word, Excel, and PowerPoint, and another tool for generic things on the internet. There are all these models. It’s like, which model do we use? And to turn that loose on people is unlikely to work well. You have to be thoughtful about it, but to be thoughtful about it, you have to understand it and how to use it. It is a two-phased approach: what can it do, and what are other people using it for? Great. Given that, what do we think we want to use it for? How do we do pilots? And then, how do we start thinking about the evolution of these tools? Because a tool that works great today might not be working great relative to its peers in six to twelve months, which is unheard of in IT. The tools will be replaced so fast, and the tools are markedly different or better. I would recommend that people think about starting dedicated resources to just understanding and doing R&D with AI, if you can afford it. That’s critical.

Since we already started talking about AI and evolution, what emerging technologies and innovation opportunities should businesses in the IT industry focus on right now?

It depends on how critical AI could be for your business. In general, AI can do two things: grow top-line revenue and grow the bottom line by cutting costs.

Depending on your business, there are opportunities to use AI for both. I recommend focusing on one area first and trying to determine whether growing revenue or cutting costs is the priority, given your business and landscape. If you’ve saturated the market, have a lot of share, and customer acquisition is challenging, focusing on cost reduction might be a better approach. Trying to focus on both with AI will be challenging because they require very different technologies.

Once you have a strategy, figure out where the costs are in the business or where the opportunities for growth are in your industry, and then try to find tools that will support your team. In that process, I should mention looking at independent consultants or traditional consulting firms who have experience with AI, because they will save you months of effort. If you’re trying to do it alone and you’ve never done this before, it’s very challenging. I wouldn’t want to climb Mount Everest with people who haven’t been there. That’s insane. Yet people often go into new ventures without experienced guidance. If you can avoid going on Mount Everest without someone who’s been there before, without an experienced guide, you really should find someone who’s done it because the amount of learning is almost endless, and it changes every few months. Trying to understand the last couple of years and learn what’s happening in the coming months is very challenging.

Ideally, you could have someone at your disposal who understands the history and the incremental changes happening, putting you in a good position for quick success. Otherwise, it could be another failed attempt. If you look at the statistics of what happened with AI initially, many products went into production, and then companies realised they weren’t working as hoped, leading to those projects being shut down. We went backwards as an industry, specifically in AI technology. Today, fewer things are in production than there were two years ago because people realise there are challenges they hadn’t anticipated. Any learning you can get from someone else is crucial for your chances of success with the technology.

How do you balance the need for innovation with the demand for stability and security in your IT systems?

One thing that many people don’t understand about AI is its probabilistic nature. If you ask an AI the same question ten times, it might give you the same answer eight times and slightly different answers the other two times. This is a very different way of dealing with technology. We’re used to a consistent output for a given input. The number one thing companies need to do when using AI is to build robust testing harnesses, which means really testing and validating the inputs to the outputs so you can understand how it works over time. Without that, it will be hard to determine if AI is working or giving the right answers. If you have no way of testing it over time, you have no way of knowing if that’s true.

Companies, in general, are not focused enough on quality. There’s a question I ask when talking to business or technology folks: How do you know if your systems are functioning right now? If you need to talk to someone to answer that, you’re behind on quality because customers can call you and say your systems aren’t working, which is not ideal. How do you start moving the organisation to a place where things are working, and if they aren’t, you get an alert? It’s the same with AI. If you keep asking the AI the same questions and it’s not working, we should say, “Hey, this thing is giving really weird results; something has changed; we need to figure out what it is.” That requires a big effort on quality, and in my opinion, very few companies do a great job with quality.

Is it a lot of testing and trial-and-error methods?

Mostly ongoing testing, all the time. Prioritising testing needs to be a fundamental part of what you’re doing because it’s no longer a case of “Oh, I ran it today, and it worked fine. Then, tomorrow it won’t work fine.” You need to be able to check that because if people are making business decisions based on information from systems, that information needs to be reliable. If it’s not reliable, which is why a lot of those production systems got taken offline, it’s because they realised, “Oh, it’s not as reliable as people expect it to be,” and that causes a lot of problems. Certainly, the worst case is making a different decision based on bad data. The best case is that people don’t trust the technology, and it casts a shadow over all technology from a user perspective, which is also very bad for the business because once people lose trust in the technology, it’s very challenging to regain it.

Reflecting on your experience with knowledge-sharing discussions, what’s been the most memorable insight or takeaway you apply in your work?

One of the reasons I enjoy Ortus knowledge-sharing events is that you meet a lot of different people who’ve had a lot of different experiences, and they’re typically curious. There are two kinds of curiosity. One is the curiosity of wondering why something is or if it could be better or different, leading to action. I’d say that’s a small percentage of the population, maybe 10% or 20%, but not 90%.

Then other people wonder what’s going on and might find out one day. That’s the larger group. Ortus has a good mix of both because you need someone to ask questions and be curious, and then you need another group that’s going out there, getting the answers, doing the experimentation, and sharing their results. These kinds of events help me understand how other people are looking at similar problems, situations that might be a little different from my world, but very applicable to it. I enjoy that aspect of understanding how people are thinking about things, especially now with things like AI, which are changing so fast.

By definition, no one is an expert. No one has decades of experience in LLMs and how they function. Certainly, people have experience with AI, and that experience is helpful, but there are still a lot of unknowns. You have to go out there, talk to people, and find out who’s doing what to understand how to deal with this.

Also, you gain resources. Plenty of people can point you to resources and things you don’t know about because they’re brand new. It is such a different world. I actually can’t believe how lucky I am to be alive right now and able to see this develop in real time with more experience.

I saw the dot-com era, the rise of social media, and the rise of crypto, and I didn’t necessarily have enough experience at that time to understand what it meant or how I could use it to my and my company’s benefit. Now, being older, it’s pretty amazing once you have the business experience and get to watch something like this happening in real time. It gives you a sense of, “Wow, okay, I’m at the forefront, and I can use this to help a lot of different people.” That’s pretty exciting. And to share that early on, people who come to these events also share that excitement. It’s nice to find peers with whom you share at least a passion for the technology, a passion for learning, and a curiosity about how to do things better.

For someone attending an Ortus Club event for the first time, what advice would you give them to get the most value from their experience?

I would say come with an open mind and no expectations. 

I’ve been to two or three of these events over the last six years or so. It was a long time ago, pre-COVID. It was my first one. I wasn’t sure what to expect, so I didn’t have any expectations, and all the events are different. You can’t say, “Oh, this is how all the events are,” because it just depends on the people, and the people are very different. The time, the topic, and everything else are very different. If you’re looking to get something specific out of it, you’ll probably be disappointed because it varies so much. However, if you’re just curious and go in with an open mind, and you think about how what someone just said applies to what you’re doing, what’s nice is that it’s a small enough group where you can ask questions. You can just say, “Hey, how does that apply to this, or does that apply at all?” Typically, the speakers have a good amount of experience in a very specific area, and they can say, “Oh, yeah, here’s how it might apply,” or “I don’t think it applies at all.” And you’re like, “OK, great. Now, I don’t have to waste my time researching that. I’m getting that really good, low-level information from someone who knows what they’re talking about.” Having that open mind is critical because if you go into it thinking, “Oh, I’m going to learn a lot about this one very specific topic,” maybe you will, maybe you won’t. There were some conversations; the last one I went to, I thought, “Hmm, I don’t think I’m going to learn much from this,” but maybe I was wrong. And it was the best event I’ve ever been to. It’s critical that you keep that open mind and don’t have specific expectations.

Lastly, what’s one question all CTOs should be asking themselves today?

The question people should be asking themselves is: What are the critical business and technology imperatives for my business that are critical in one to two years? For most executives, that is a very short time frame. One to two years is no time at all, frankly; roadmaps of five years or three years are not as interesting.

I was recently talking to an AI company that’s helping with some quality aspects around AI, and I asked them where this was on their roadmap. They said it was on there somewhere, but they just didn’t know the specifics because their roadmap is about three months, sometimes six months out. I thought it was crazy for a product company selling to the enterprise to have such a short roadmap. They explained that the space is evolving so rapidly, they can’t plan longer than that because they just don’t know what’s going to come and disrupt what they’re doing. So they’re looking at very short, incremental cycles with maybe one big investment. I thought, “Wow, that’s different,” because typically, in prior experiences, it was really about knowing where we are today and roughly where we need to be in three or five years, and knowing the crawl, walk, run path to get there. Now, it’s like, I may not need to walk or run; actually, I may not need to run in three to five years; I may need to fly or swim. And that is just a very different animal from running. How do I figure that out? And how do I stay lean? So I’m thinking about maybe this year’s evolution and maybe one other revolution, and then that’s it.

I would say shrink planning cycles, shrink roadmap cycles, and try to focus and figure out where the business and technology are going in the next few years, because I’m seeing massive and unprecedented disruptions to businesses with the help of technology. You need to be thinking about that for yourself and your business.