3163: Teradata CTO on the Evolution of AI, Personalization, and Data

[00:00:03] So, what can we expect this year for AI, data, and enterprise technology? We hear about all those things on our newsfeed, but what does it really mean? Well, my guest today, Lewis Landry, is the CTO of Teradata, and together, we're going to be exploring the evolving landscape of AI, personalization, data analytics, and trusted AI.

[00:00:26] With more than two decades of experience in software architecture and engineering leadership, my guest today has played a pivotal role in shaping how businesses harness data for intelligent decision-making. So, we'll discuss the next wave of AI advancements from Retrieval Augmented Generation, or RAG, to agentic AI systems that seamlessly blend generative models with traditional software.

[00:00:55] And my guest will also break down why trusted AI is essential for businesses, emphasizing transparency, accountability, and measurable value creation. So, with AI continuing to transform how enterprises operate, one of the big questions out there is, is your business truly prepared for what's next? And how can you help build AI systems that enhance, not erode trust?

[00:01:23] And with that seen perfectly sound, it's time to introduce you to today's guest. So, thanks for joining me on the podcast today, Lewis. Can you tell everyone listening a little about who you are and what you do? Absolutely. Thanks for having me. I am the newly appointed CTO at Teradata, where we're focused on empowering organizations with trusted information and intelligence.

[00:01:47] At Teradata, we believe that people thrive when empowered with better data insights and our technology aims to make that possible. As CTO, I lead our technological vision and innovation strategy. Focused a lot on our open and connected platform capabilities.

[00:02:03] And, you know, as is everybody today, we are also focused on bringing next generation AI driven value to our customers and making sure that they can get that operating in whatever environment works best for their needs. Before this, I had several leadership positions at Teradata, almost entirely working on emerging technologies. And before that, worked at a couple of other major technology companies.

[00:02:33] And honestly, I nerd out so hard on the tech that it's exciting to be able to be in a role like this, helping to drive technology innovation in a company kind of at the epicenter of data and AI. Well, thank you for joining me on the podcast today. And we booked you in some time ago on this podcast. I think it was when you very first got promoted to the role of CTO. So first of all, a belated congrats from me.

[00:03:02] But I'm curious, how were your previous experiences as head of technology and innovation office? How has that helped shaped your vision in this relatively new role? The thing about it is, through my tenure at Teradata, I spent most of my time looking at, like I said, emerging technologies and new things. Being able to lead that group, which is an advanced research and development group, allowed me to keep my eye on what's coming.

[00:03:29] A lot of times when you're working in enterprise companies or companies that serve enterprise companies, you're very focused on the here and now and maybe what's immediately next. We got to spend a lot of time thinking about what's possible.

[00:03:47] Exploring things like how to transform our existing engine into a different configuration or a different form factor to be able to do a much more on demand and serverless type workloads.

[00:04:01] As well as things in data discovery and data relationship discovery, which led to some of the early AI work that we did in what we call Teradata's Ask.ai, which is really about putting a conversational interface on top of the systems that we offer and obviously the data that they manage. And we are pretty much more in Q1 still of 2025.

[00:04:30] It is still early days. So I've got to ask, from all the tech trends that we're seeing right off the bat, we had CES recently. There's so many big announcements there. But what are the key trends and advancements in tech that excite you or that you think will define the rest of the year, particularly in areas like large scale personalization, enterprise retrieval, augmented generation, all that cool stuff that we're hearing about at the moment? Yeah, you hit the nail on the head.

[00:04:57] The very first thing that I think is incredibly interesting is a maturity and retrieval augmented generation. You know, this isn't a new concept, though for a lot of people it still is. But I think what we're going to see is a really large maturity in that, especially for large enterprises.

[00:05:18] It's like things go from little sort of toy implementations and experimental things into real production and mission critical systems. And a big part of that is making sure that we treat vector embedding data and vector stores like enterprise systems. And that's something that's really, really exciting for us. Another thing around personalization.

[00:05:46] I think that one of the coolest things about this wave in general of AI is that it offers the promise to be able to give people personalized experiences. It offers an ability to sort of mold to how you communicate, how you interact with technology. And that's something that I suppose has been possible in the past, but not nearly like we're seeing today. So very excited about seeing where that goes.

[00:06:17] I would be remiss if I didn't mention that much like we've seen over the last 10, 15 years in computer vision, where if you imagine 10, 15 years ago, how expensive it was for Google to figure out whether there was a cat in a picture. To now, you can buy a relatively cheap camera with that technology built into it that runs on a tiny little chip in there for next to nothing. And I think we're going to see the same kind of evolution.

[00:06:44] We're already starting to see it with generative AI models where, you know, it used to take a farm of GPUs to do some very, very cool things. But now these models are getting tighter through practices like distillation and a handful of other techniques to make it much cheaper to do very, very cool things. So that's really exciting. And last but not least, I am most excited, I think, for this shift towards agentic AI systems.

[00:07:13] In part because, to me, it really starts to show how we can blend the aspects of generative AI with traditional software to build really autonomous systems

[00:07:27] that react and handle the diversity of inputs and outputs and handle the non-deterministic realities of the systems that we build. So I'm really excited about that. And it's something that we're very focused on as well. And one of the other things that really stood out to me about your approach at Teradata is it's less about the hype and shiny new stuff,

[00:07:57] but it's much more about building trusted AI. So on behalf of every business leader listening, you'll probably find that incredibly refreshing approach too. What steps do you think their organization should be taking to build trust in AI? And why is it so critical for businesses moving forward? Yeah. Also, great. This is central to who we are, actually. Trust is something that I learned very early on in my career.

[00:08:27] Trust in data is central to everything. Anybody who's in the data industry needs to be thinking about that. Because if you can't trust what you're seeing, then you can't make reasonable decisions based on it. So for us, when we think about trusted AI, we think about sort of three fundamental principles. One is people-centered accountability.

[00:08:52] Making sure that AI is augmenting human capabilities rather than completely replacing human judgment. Now, we know that with these autonomous systems and things like that, especially the rise of agentic AI, a lot of decisions are going to happen through autonomous loops. But it's really, really important that the accountability remains centered on people. And that brings us to a second, which is transparency.

[00:09:20] Being able to understand how and why an AI-driven decision was made. And it's not just about technical transparency. It's about making these decisions interpretable and explainable to any stakeholder at any level. And last, but certainly not least, is around value creation. If what the AI is doing is not driving value, it really doesn't matter. And it has to be able to deliver measurable business impact while maintaining trust.

[00:09:49] And for our customers at enterprise scale. It's, for me, really important to understand that the trust accelerates the opportunity. And without the underlying trusted data, honestly, none of these investments pay off. They turn into just really cool demos.

[00:10:09] And again, for us, it's really about making sure that our customers can innovate quickly, but also responsibly. Because they can trust that the data at the base layer, the foundational layer, is exactly what they need to get the right results out of the AI systems that they build. And as an ex-IT guy, race on the belts and braces approach of you can only improve what you measure.

[00:10:37] Again, incredibly refreshing to hear. And adding on to that, I mean, with your extensive experience in analytics and data platforms, how do you see the role of data evolving in that decision-making process and, indeed, innovation this year? Yeah, so evolving is an interesting way to frame that. Like, data is at the foundation of decision-making today.

[00:11:04] But I think what I'm seeing is, again, with the rise of this kind of AI, we have a new type of data that we have to go treat with the same rigor and care as we do this other data. In a lot of places, when we think of vector embeddings, they're in little sort of purpose-built databases, or they're kind of like off to the side of a data management practice or a data management process.

[00:11:31] And, you know, that's fine for the pre-Cambian explosion that we're seeing in discovery and that kind of stuff. But I think as 2025 continues, we see more focus on trying to not, like, rein it in as far as slow it down, but start to really manage and govern that and take it more seriously across the enterprise. I think beyond that, high-quality data is absolutely essential for making informed decisions,

[00:12:00] enhancing collaboration, conducting accurate analysis, and ensuring trustworthiness in any system or process. What's interesting is the role of data is evolving, but mostly I think it's really that the data that we're using is evolving. And I'm curious, as someone with a background in designing real-time data acquisition platforms and analytics as a service,

[00:12:29] how's that influenced your approach to driving technological innovation at Teradata? I would imagine it must play quite a big part. Yeah, it definitely is influenced the way that I think about things. Working in real-time data acquisition taught me a lot of lessons around making sure that things that we capture are actionable, that they integrate really, really quickly.

[00:13:01] And also, I think more importantly, and this really ties into Teradata, building the importance of building really scalable, reliable foundations that can support both rapid innovation, but also that can support a very bursty kind of workload, that can support really anything that you throw at it. There's not a lot of room for error in those sorts of systems,

[00:13:26] and especially when you're depending upon fast turnaround on results. The latency matters, the reliability matters, and that goes directly into how we think about building everything that we do at Teradata. You've got the air of so many different businesses in multiple industries out there. So I'm curious, from all those conversations that you're having, what challenges do you think that businesses are facing right now

[00:13:54] in adopting new technologies in 2025, and how can they better prepare for some of these changes that you must be hearing about and talking about on an almost daily basis right now? There's a lot. One of the interesting things that I initially had not thought a lot about is the increasing concern around AI regulation at local and national levels.

[00:14:22] This landscape will obviously continue to evolve, and obviously it's diverse depending upon where you are. But it really is causing a lot of folks to not necessarily pause in what they're doing, but have to in parallel think pretty hard about how they maintain governable foundations, how they maintain the ability to explain autonomous systems and build on these trusted foundations.

[00:14:53] That can be really challenging, and it's definitely on the minds of a lot of leaders that I've spent time with. Secondly, there's a pretty big shift now from what I might characterize as project-focused AI development or AI system development to value-focused. And a lot of that is because we spent the last couple of years following what might be the most impressive technology demonstration

[00:15:22] in the history of humanity with ChatGPT, working through how this is going to actually, how it's going to affect the business processes that we have, how it's going to enable new types of workflows, and all the while learning this technology's limits, learning its capabilities, and being surprised, it feels like, every other week with a new research paper that comes out.

[00:15:49] And so shifting from just sort of like raw investment and exploration projects and things like that into real value-focused implementations is a challenge. In some part because things are still moving so fast. It's hard for us technologists to keep our eyes on business objectives when the shiny objects are flying by so fast.

[00:16:20] I think the agentic systems are going to help a lot with this. And making sure that we're balancing the investment against returns is also really important, especially in the economic climate that we're in. And, you know, ultimately, trying to evaluate not all of the projects in this space have borne fruit over the last couple of years. So taking a more rigorous approach to that, I think, is in a lot of people's minds

[00:16:51] and something that we're definitely trying to help with as much as possible. And again, in all of this, it comes back to starting with clear business outcomes and objectives and then building on trusted data foundations to make sure that you can actually achieve those goals and not assume that everything just gets magicked into the outcomes that you're looking for. And again, music to my ears hearing you talking around business outcomes, business objectives,

[00:17:21] rather than just the technology. When I was researching you before you came on, I was reading how Teradata emphasizes harmonized data and cloud analytics for AI. But how does this approach enable organizations to meet some of those business outcomes and objectives to achieve faster innovation and more impactful business results? Because that's the goal that everybody's chasing, isn't it? And I think very often we get distracted by the tech itself. But anything you can share around that? Yeah.

[00:17:49] So data harmonization or data integration is one of the founding strengths of our approach. Teradata has been, I would argue, the leader in this for decades. And harmonization is fundamental to enabling innovation and achieving better business results. Like I said, it's something that we've seen over decades. At a high level, the data we work with is a reflection of the business processes and systems that exist within an organization.

[00:18:19] You could think of that data as exhaust from those processes in some cases, as side effects of it. But really, it's a reflection of everything that's going on. Those business processes and systems, they don't exist in isolation. They have dependencies on each other. They impact each other for certain. And in a lot of ways, they are an interconnected fabric of people

[00:18:46] and activities and tasks and decisions. So trying to analyze and explain them in isolation is challenging, if even possible. And bringing all of that data together in a way that is accessible and integrated allows for more effective data analysis, confident decision-making, and improved strategic planning. It's a lot like going to see a symphony performance

[00:19:15] where, as you may expect, so much better if the strings are in harmony with percussion, brass, and woodwinds. Just looking at it through one of those sections or listening to one of those sections, while good, isn't going to be nearly as good as the entire symphony in concert. And what makes this particularly powerful in today's environment is how it enables organizations to leverage AI and analytics across their entire data state while maintaining trust and guards.

[00:19:44] And as someone that I was reading has contributed significantly to the open source community throughout your career, I'm curious. How do you see the role of open source technologies evolving in enterprise environments? And are there any opportunities that you think this presents for greater innovation as well? What are you seeing here? Yeah. So I am a big believer in open source. I spent my early career almost entirely working in and around open source, both in development and also in leadership.

[00:20:14] And I think this really shaped me a lot. One particular thing in that space for me personally is getting to collaborate with and lead a pretty broad and diverse set of people across regions and cultures. For someone in their 20s trying to figure out a career path, this was an incredibly valuable thing for me and has really kind of helped me on my journey,

[00:20:42] learning a lot through early career trial and error on the people side of things. As to enterprise environments, I mean, it's probably fair to say that most enterprise technology teams are dependent on a lot of open source tech today. And one of the ways that I see it used is in quickly diffusing tech through the global community. Let's use LLMs, for example. Now, while not alone, I think we can thank Meta and the teams working on Llama

[00:21:11] for a lot of how quick and diverse technologists have been able to push innovation in LLM technology. Same can be said for things, the folks behind LengChain and related things. I don't think without that open source push, we'd have the same accessibility, cost effectiveness, nor the same flexibility and customization that enterprises thrive on with that kind of technology. Ultimately, the open source community

[00:21:42] pushes a lot of technology forward that allows us all to build our more bespoke or more purpose-focused solutions. And that's incredible. Being able to choose the technology you own versus outsourcing everything to a SaaS provider harkens back to this concept of trust that Teradata cares about a lot as well. Maybe that trust is around privacy and sovereignty, and maybe it's just about trust and cost management.

[00:22:10] But it's certainly something we concern ourselves with, and I'm just a huge believer that the open source community and the work that goes into that pushes all of us forward and, as I said, helps really just sort of diffuse new things into the overall global technology community. Well, thank you so much for sitting down with me today and sharing your insights from two decades of experience

[00:22:37] in software architecture and engineering leadership and driving a company's technological vision and innovation strategy. But it's time to have a little bit of fun with you now and ask you to also leave my guests with two tiny little gifts. Yeah, it's that time of the podcast where I'm going to ask you to leave either a book that means something to you or that we can add to our Amazon wishlist and a song that we can add to our Spotify playlist. I don't mind what you leave everyone listening with,

[00:23:04] but what would you like to leave and why? If you'll allow me, I'll give you one of each. Oh, yes. I would love to recommend The Innovators by Walter Isaacson as a book. As someone who studied both history and computer science, this book resonates deeply with me. It traces the evolution of computing from Babbage and Lovelace through to modern innovations, revealing how combining theory with practice drives real transformation.

[00:23:34] That's something that stuck with me since I read it originally. And it shapes the way that I think about how we execute on innovation and discovery. For the playlist, I scanned through what you've got and it's a pretty incredible list. I would add the song Don't Carry It All by the Decembrists. Over the last year, it's been a powerful reminder that especially in times of great change, innovation and progress are team efforts.

[00:24:03] The theme of the song hits hard from the very first line. Here we come to a turning of the season, which really resonates for me in the huge shift that we're seeing in AI technologies in our market today. And later on talks about how we must all carry each other's burdens and drive towards our goals together. Obviously, it's talking more about a harvest when you listen to the song.

[00:24:30] But to some degree, isn't that what we're trying to do as techies as well? Harvesting value out of the seeds and toils of our work. I also do the song an awful lot. Yeah, great band too. There's a few tracks there as I used to love. It was We Both Go Down Together, The Crane Wives 3. Oh, yes. So that will definitely go on the Spotify playlist. And I'll also add the great book as well to Amazon wishlist. And for anyone listening wanting to find out more information about Teradata,

[00:25:00] dig a little bit deeper on anything we talked about today. Where would you like to point them? Yeah, so everybody can go straight to Teradata's website at teradata.com to learn more about the Teradata solutions, the trusted AI capabilities we have. We also have a Medium blog. Our team is always excited to discuss how we can help any organization navigate the future of enterprise AI analytics and data.

[00:25:28] Well, we've covered so much today from what the tech trends will dominate 2025, what businesses should expect or be aware of, and indeed most importantly, of all the importance of trusted AI, and even a chance to leave me with a great book and a cracking song too. So thank you for joining me today, Lewis. Thank you for having me on. I think it's clear that trust in AI is equally as important as the technology itself. Whether it be the retrieval augmented generation

[00:25:58] or RAG or agentic AI systems or to data harmonization and cloud analytics, the future of enterprise technology is being shaped by how well businesses integrate AI while also maintaining transparency and accountability. Because AI isn't just about automation or efficiency, it's also about empowering better decision-making with high-quality trusted data.

[00:26:26] And as companies navigate challenges of AI regulation, governance and investment, the real winners are going to be those that build AI systems that enhance, not replace human expertise. But what are your thoughts on the future of AI and enterprise environments? Are businesses moving fast enough? Or are they struggling to keep up with the pace of innovation? I suspect it's somewhere in between, but let me know your perspective.

[00:26:56] Techblogwriteroutlook.com LinkedIn, X, Instagram, at Neil C. Hughes. Easy to find. But don't just hit follow. Please say hello. But it's time for me to go now. So keep exploring how technology is reshaping the way we work, innovate and build trust in AI. And don't forget to meet me back here tomorrow because I've got another cracking guest lined up. But thank you for listening as always. I'll speak with you all again tomorrow. Bye for now.