Does Vendor Lock-In Belong in an Enterprise AI Data Platform?

Speaker A: Foreign. Welcome to the Don't Panic It's Just Data podcast and we are recording live from Big Data London and we are on the floor today. And here we have with us, uh, Terry and Andy.

Speaker B: Hi everyone. So My name's Terry McCann, I'm um, the CEO of Advancing analytics, the lovely uh, booth that you can kind of see behind you here. Um, and yeah, great to be, ah, here. Hi everyone. Uh, Andrew Fithian. I lead our UK enterprise growth team at uh, Databricks.

Speaker A: Great to have you both on the podcast and we're so excited here to be on the floor interviewing you. How has Big Data London been for you?

Speaker B: I love this event. Right. I always think, um, you know, we, we get the pleasure of doing many, many different events across the year and um, Big Data London has always been one of my favorites. You know, I've kind of attended this event as, you know, both an attendee, as a sponsor, as a speaker for the last kind of 10 years. And um, and I always think it's a great opportunity, you know, not only to kind of see like our great partners like databricks, but also over those kind of 10 years you get to know a huge amount of people, you know, from people who have worked for us, people who've worked for competition, you know, customers who have gone off to join many different companies. You know, it's almost uh, it's almost a kind of homecoming of lots of different people and you forever just, you know, bump into people you might not have seen for a couple of years, pick up those kinds of conversations and those relationships and um, you know, I think it's always great for that and uh, a great opportunity to kind of meet new people this year that hopefully again, we'll see throughout the rest of the years, kind of, you know, continuing with the event. So ah, I always love being here.

Speaker A: What about you, Andy?

Speaker C: Yeah, absolutely love the event. I mean this is, this is actually my second one so far. So like having, coming last year as a newbie and sort of experiencing it, I completely emphasize, you know, the, the familiar faces I've already seen this year, the energy that's here. Also the change has been phenomenal. What we were talking about last year is completely different to what's going on right now. And it's a real industry event. It's not about one vendor, one area. It's actually a real industry perspective of the actual data and AI industry as a whole, which is really interesting when you think about it from that perspective.

Speaker A: So this is actually my third Big Data London. And when I came here a bit over two years ago now. It was all about gen generative AI, then it was about agent AI last year, and this year it's all about context. So what is the one key, most exciting thing that you've seen on the floor this year? Andy, why don't you go. Yeah,

Speaker C: yeah, Look, I think for me, uh, from a databridge perspective, the. Like you say, there's a lot of buzzwords going around. I think, like, context is so important when we're looking at, when we're looking at generative AI, especially as we're seeing customers trying to not just think about experimentation anymore. It's actually about execution of that, but the way that they grasp it and the way that they're looking at, uh, the answers. And engaging with vendors to try and figure out that to drive true business outcome is super exciting for me. Uh, especially the conversations we're having and everything else when it goes with that.

Speaker B: Yeah, for sure. We spend a lot of time with our customers thinking about ontology and thinking about, I think, I think the shift away from, you know, talking about generative AI to then talking about kind of everyone's talking about governance last year. And I think, you know, when you look at those two things combined is where we end up, you know, thinking about, you know, ontology and, you know, thinking about how we govern data, uh, almost less for the individuals in our business and now more for the agents who are working with us. And so there's been a, you know, definitely a huge focus around, well, how are we, how are we really thinking about adding that additional layer of context on what we're doing, you know, which is, again, why, with things like, you know, GENIE from databricks and looking at kind of ontology, there is a, A real push around, you know, how do we bring this to life to make sure that the decisions that, uh, you know, the questions that we're asking to our models is ultimately, you know, providing the right information back, you know, in the context of our business, with the understanding of our business. And so that we're not, you know, ending up with just garbage in, garbage out, that we're getting, you know, the best answers we possibly can do.

Speaker A: What is the biggest challenge that you are facing still and trying to kind of tackle?

Speaker B: Yeah. So, uh, I would love to say, I think, you know, that every customer is, you know, AI ready. Right. But I think the reality is, you know, I think a lot of people think that you can, you know, you can go and you can buy anthropic and you can go and buy OpenAI and you can put it on top of your data and you'll just get the right results, right. But it becomes, you know, you become unstuck very quick without having, you know, a great platform, you know, like something like databricks, where, you know, you have a truly kind of governed data platform that has been well modeled, you know, and is in the kind of understanding and the language of how your business speaks. And so I think, you know, when we, a lot of the people we talk to, you know, their businesses have gone and you know, bought um, you know, an anthropic or something similar and then they haven't really considered, well, how do we do the plumbing, how do we get, you know, the data right, that really enables what we're doing, you know, And I think that's one of the most common kind of stories that we're hearing and how we're trying to help people position with, you know, whilst we might have again done data models to uh, help a business answer a question, now it's very much how are we doing that to really power, you know, great things like genie so that we can, you know, enable conversational intelligence across all data.

Speaker A: What about you? What is the biggest challenge that you have seen?

Speaker C: Look, I mean as a whole, I think it's that shift into execution. How do you drive genuine business value from agentic systems? I think some of the challenges we see are kind of fourfold. We call them sort of the four Cs, so context being one of them. Absolutely. But also control. How do you put the right governance across all of these different layers? How do you ensure that the cost as well is actually manageable? You're not token maxing anymore. Uh, in an aspect of actually how do you value max? How do you create the value from the organization? And then finally choice, you know, like, you know, vendor lock in is not a strategy, it's how do you actually have the choice across the different vendors, across the different cloud areas, across the different LLM components. All of that is really super key. So we sort of tackle that as those challenges of the four C's in databricks.

Speaker A: I think I recently did a podcast on the four Cs and one of them which is also keeping your data current. So I think I would add that now it's like 5Cs that you need to.

Speaker C: Well, uh, I think you've sort of touched on another area there which is almost sort of like real time analytics. So we've just, we've just released um, our own new uh, products, you know, code name Raiden. And that is really around how quickly can you get context from data that's actually being put in? And actually that is a super important aspect of I guess, what organizations can drive value in from the future in terms of how quickly they can turn inputs into outputs that, that real time analytics is super important.

Speaker A: Is that why Advancing analytics has partnered with databricks? What are some of the most, most key benefits of the partnership?

Speaker B: Yeah, for sure. I mean, you know, we've been a databricks partner for pretty much about seven years now. And so I have had the kind of pleasure of seeing the platform really change, you know, and like when we look at, when we look at the sort of problems that our customers solve, you know, we partner uh, you know, with businesses who have, you know, genuine problems, who have big data and are looking, you know, to solve something in the best way they possibly can do, you know, and that's why if I look at the kind of, you know, trajectory that we've been on as well in partnership with databricks. So you, um, know, we were really pleased very recently to become a databricks venture portfolio company. And so databricks kind of co invested alongside our primary investors, um, in us because you know, we all are aligned in terms of the vision that we're trying to get across, which is how do we bring value to customers in the best way that we can do, you know, in a way that is both cost effective, but really, you know, to Andy's point, helps them drive and helps them deliver those outcomes. And so when it comes to, you know, what is it that we're looking at? I think when you look at, when you look across big data, London, right, There are a lot of tools where you could say, okay, well these look and feel similar. But it's only when you really look at, you know, a lot of these tools are solving a singular problem. Whereas what databricks is really trying to do is to be, you know, the data ecosystem. So it's not just a, uh, data platform, an agentic platform, you know, we're looking at, well, how do we solve things like entity resolution for Customer360, how do we solve, you know, security and SIEM? How do we, you know, think about deprecating, you, uh, know, OLTP databases and running everything under one ecosystem. And I think when our customers look for a platform, they don't just want to be like, well, I need to go and buy 15 different tools. You know, they want to make sure that they're committed to one platform, you know, that they know is going to go places, that is continually invested in, you know, that is backed by the best team. And ultimately, you know, that's a uh, conviction that we really believe in, that we see really, you know, pay dividends in our, in our customers. For sure.

Speaker A: Coming back to the vision that you talked about it being aligned with databricks, can you give us a customer story, post the investment?

Speaker B: Yeah, for sure. So I always can't give names. Right. But um, one of the things that you know, we've been doing recently is part of, you know, part of the investment, you know, and where we're ultimately looking to kind of co invest is how we go about building out the tools, you know, that enable our customers to take advantage of databricks, you know, in the fastest way possible. And so in combination with databricks and with GENIE code, you know, we came to market with a piece of IP that we call Lake Forge. And Lake Forge is ultimately an agentic platform for building out databricks, um, lakeflow patterns as fast as possible. And so we've been deploying that with um, one of our hospitality customers, one of the largest kind of, um, food and beverage customers, uh, in the world. And what they've been really looking to do is they have huge amounts of data, massive amounts of acquisitions that they go through, you know, global kind of coverage. And the difficulty they have is, you know, so many different systems, so many disparate sources and what they need is they just need fast and they need reliable and they need repeatable. And so what we've been able to do is really help them drive, you know, the ability to say, well, rather than it taking, you know, weeks or months to onboard a data source, you know, through working through Lake Forge and GENIE code, you know, bringing that down into minutes so that they're able to get access to, you know, position that data back out to their business units to work out, you know, where they're selling, you know, where they're hemorrhaging margin and ultimately, you know, deliver a much more, you know, business focused outcome for their company.

Speaker A: So Andy, I have a question for you.

Speaker C: Sure.

Speaker A: What's the biggest shift in how businesses use Data, uh, that CEOs aren't paying enough attention to right now?

Speaker C: I would have normally gone down the context route because, because it's a bit of a buzzword here, but I think it's never truer in that sort of sense. Actually it's not necessary that AI isn't. AI is smart enough. Today, the new Models that uh, they're bringing out. We don't necessarily need them for all different business operations and things like that. So for us it's really around how do you scale that, that motion where you can leverage, whether it be open source models, whether it be lower cost models, whether it be machine learning, leverage your foundational data at the bottom to then bring out the, bring out more value in what you're trying to achieve from uh, a business perspective. I think a lot of time organizations uh, are experimenting with the big vendors of the world and putting a lot of investment in that but not necessarily putting that into, into production properly. Um, the other thing I think as well is when I think about it more, it's, it's almost how do you, how do you set your organization up, uh, to be putting the investments into the revenue generating areas of, of what you're trying to achieve and not just the cost savings, not just the efficiency drivers. We already seeing that now. But actually when we're thinking about that from a, uh, you know, the, what are the big ticket items that are really going to move the needle? Especially with industries changing models changing, that's super important. I think it's almost that innovator's dilemma in the sense of how do you, how do you go from what is normal now? You can't just improve that with AI, you kind of got to redefine it in that sort of sense.

Speaker A: I was just going to come to that. When you said innovators dilemma, I think you really caught that. So my follow up question to that is because CEOs are quite focused on what serves uh, your ROI the best, right? But they may not see how employees need to leverage some of these other tools or like something in the company that may not be generating enough roi. So what is it that they need to turn their attention to despite it not bringing up the numbers?

Speaker C: A large question. Again, I think if I sort of try and simplify it from uh, from our own lens, I think there's a lot of value and untapped value in, in a lot of the, the data that organizations have today, whether it be structured or unstructured, whether they have access to it today or or not. I think it's putting those foundations and fundamentals in place first and foremost so that the gift then gives them the ability to sort of leverage that and start to do some of m. Those more bleeding edge practices that we talked about today, that that would be mine and do that with good, you know, do that with good partnership. You know, you look at A, uh, partner of ours, like advancing analytics, they have the expertise, whether that be within our own, um, partner ecosystem. They're also practitioners. They're not just advisors. They're really in it for the, you know, they're in the right. On the, on the journey with us in that sort of sense. Um, and they share that same vision as us. So having that, having that ability to partner with people like that will really help the success and drive the outcomes that we need to for the organizations.

Speaker A: Terry, is there something that you have come to a realization of as CEO?

Speaker B: Yeah, I think if I look at, you know, we often talk a lot about, you know, justifying the ROI of an investment into data or an investment into AI, But I also think, like, do you know what? There are companies that I meet at Big Data London year after year after year after year, and part of the problem isn't necessarily justifying the roi. Like, it's a lot of it about trying to understand, you know, actually the kind of cost of inaction and that kind of, you know, when you look at some of these companies who come, you know, and they talk, but they don't actually commit, and you see their businesses not growing, you, uh, know, in comparison to their competition. And so a lot of it is really about, you know, actually, actually moving and committing. And I think when you, when you attend an event like this, you can kind of be a little bit, you know, blind to the fact that you think everybody is, you know, a very progressive business. But. And the reality is there's a lot of people who, you know, are still a little bit too hung up in potentially, you know, what might not happen. And, um, you know, really that kind of cost of inaction is holding those businesses back, you know, and the more and more you do it, the more that they're not investing in platforms like databricks, you know, the more that they'll find that, you know, their competition are able to get to market faster. And in a way that, you know, they better understand their customers and they better understand the outcomes they're driving. And that, uh, you know, they'll find that, uh, you know, not only will their customers leave, their employees will start, you know, not wanting to work for such a, you know, a business that isn't investing in the right areas. And so I think that, you know, we talk a lot about roi, but it's that kind of like COI as well, that's super important that I think a lot of the times just kind of gets forgotten.

Speaker A: So I have two fun questions for you One is a more industry centric. So what is the one thing that the tech industry is getting completely wrong about is AI today?

Speaker B: Yeah, um, I think it goes back to the idea, you know, that the boardroom can make a commitment to a, uh, you know, to a foundation model company and then they will just assume it will happen, the outcomes will come. But if you don't put the investment in, you know, you're not building out the right platform, you're not thinking also about how you scale those models. You know, one of the great things about databricks and especially with like um, the AI gateway is ultimately being able to build a model and then host it and serve it out to the business in the easiest way. And a lot of the times it's a, I will buy this thing. And then, you know, it's like step one, buy the thing, step three, profit. And people generally forget about step two. And the reality is, you know, that's where things are going to start falling apart. And so unless we're really investing in both, you know, data and AI, you know, you need to have the right kind of foundational platform as well as the foundational models.

Speaker C: You agree with all of that? I think the only thing I'd add is then the people aspects, you know, and setting up uh, the organization. So these roles, these employee roles are actually heard uh, at a board level, whether that be cdo, caio, having those types of expertise at a board level because they're going to be fundamentally how organizations are moving forward is again not just experimenting and using the latest techs. How do you pull these teams together, whatever you want to call them, tiger teams, et cetera, et cetera, so that they're tackling the biggest tasks within the organization and they're driving the value that's needed or the revenue generation that's needed as part of that.

Speaker A: So a fun question for both of you. I don't know if you have your own personal AI agents or not. Maybe you do. So if you had to assign a human characteristic to your AI, ah, agent, a personality trait to your AI agent, what would it be and why?

Speaker C: Quite like a sassy agent. You know, every time I ask it something, you know, just give me a little bit of banter back or you know, you know, make it a little bit more fun rather than, you know, interacting with it on a, on a sort of purely business basis, I think that'd be quite funny and add a little bit of uh, humor to the day. It's all for me.

Speaker B: Absolutely. I often think, um, you Know, if I look at, if I look at our social content, the thing that always lands better, you know, we can produce these incredible thought leadership pieces, you know, but the things that always land the best are, ah, when you simplify a tricky problem into a meme. And so I think, you know, the more that I can have a model that just responds to me in memes, you know, and Beckham memes, anything like that. Right. You know, then, uh, then it's, you know, it's, it's speaking my language. Right. You know, it's working for me.

Speaker A: Do you say that? I am not a meme person at all. I block everyone who sends me to.

Speaker B: Well, that's it. Right? You know, now you're just going to get them from me. So

Speaker A: cut. You. Okay, so one final question for you before we end the session today. One key takeaway for CD CDIOS.

Speaker B: Yeah,

Speaker A: yeah. One key takeaway in 60 seconds. CD.

Speaker B: Yeah. So I think for a, uh, for a CDO at the moment, really, uh, what we want everyone to understand is, you know, speed to market is the most important thing right now. And so that has to be embracing not only AI to support things like, you know, your decision making, but AI for agentic engineering. You know, the faster that you can take data from one place, that you can translate that into something that solves a problem and get that out to a decision maker in your business, the better that, you know, you're going to be able to outpace your competition.

Speaker C: For me, it's really, you know, becoming that translator between the IT and the business. To Terry's point there, like, you can do all amazing things from a technology perspective, but unless it lands and IT translates and it resonates within the business, business side of things, it kind of gets lost or it gets lost in the ether. Um, so being able to really understand the business challenges, understand the technology challenges, build the right agency models that are impactful, that aren't just experimental, that are operational, and translate that to business value. That I think that's a really, a really interesting skill, um, and thing that will drive a lot of changes.

Speaker A: Yeah, well, I'm looking forward to seeing that. Is there something that you want to talk about and I haven't asked, uh, something that you missed and you would like to cover? Anything to add?

Speaker B: No, I don't think so. I think we've managed to cover, you know, a great deal. You know, thanks very much for having us here today. You know. Right. We love being here. We'll be back at Big Data London. We've Got, you know, a load of great events coming up. So, you know, I think my next event that I'm most excited for is the one that we do jointly, so the Data and AI World Tour in London. Um, so that's the Databricks UK flagship conference. We're looking forward to kind of being there and continuing to kind of grow our partnership together.

Speaker A: When is it?

Speaker B: So it is the 11th of November, um, over at the Excel in London.

Speaker A: Excited now? Uh, how should the audience find you?

Speaker B: Yep, you can find us. Um, advancing analytics. I would always say to anybody who loves going to an event like Big Data London is, you should check out our YouTube channel. You know, we have about 400 videos on there where, you know, we just do this, right? We talk about the things that we're passionate about. You know, we talk a lot about data bricks, we talk a lot about, you know, associated technologies. You know, I'm, um, trying to bring those to life in a really simplified way. And so if you like events like this, you'd love looking at our YouTube.

Speaker A: What's the website?

Speaker B: It's, uh, advancinganalytics.co.uk.

Speaker A: brilliant. Well, thank you so much, Terry and Andy, for joining us today.

Speaker B: Thanks for having us.

Speaker A: And we are live from Big Data London. This is the Don't Panic, It's Just Data podcast. Thank you so much.