Ever wondered how data governance is transforming in the age of AI? Join us as we sit down with Felix van der Malen, the visionary co-founder and CEO of Calibra, to explore this fascinating journey. Felix takes us back to the inception of Calibra during the 2008 financial crisis and explains how this pivotal moment underscored the urgent need for organizations to master their data. From the days of regulatory compliance to the rise of self-service analytics, discover how Calibra has evolved to become a leader in data governance and intelligence. Felix sheds light on the seismic impacts of big data, the necessity of data cataloging, and the transition towards cloud infrastructure, all within the context of AI’s growing importance.
Building a data-driven culture within organizations is no small feat, and Felix provides invaluable insights into this demanding process. Learn how shifting the role of data scientists to AI stewards and fostering collaboration between technical and business stakeholders can unlock AI's full potential. Felix shares real-world examples, such as Heineken’s success in creating a common language to facilitate digital transformation. We dive into best practices for cultivating a data-first mindset, stressing the importance of clear communication and shared understanding in aligning complex organizations to maximize AI impact.
Finally, we turn our attention to the concept of data product thinking, starting from business challenges and working backwards to define the necessary data and technology. Felix discusses the crucial role of data offices in ensuring governance, quality, and accessibility to maintain relevance. He also highlights the importance of aligning data initiatives with mission-critical priorities to drive strategic value. Wrapping up our conversation, Felix offers a glimpse into Calibra's unique customer-centric culture, shaped by its European roots and expanded to New York, underscoring their commitment to customer intimacy and a humble yet passionate team environment. Don't miss this episode if you're keen on understanding how to harness data governance for AI success!
Ever wondered how data governance is transforming in the age of AI? Join us as we sit down with Felix van der Malen, the visionary co-founder and CEO of Calibra, to explore this fascinating journey. Felix takes us back to the inception of Calibra during the 2008 financial crisis and explains how this pivotal moment underscored the urgent need for organizations to master their data. From the days of regulatory compliance to the rise of self-service analytics, discover how Calibra has evolved to become a leader in data governance and intelligence. Felix sheds light on the seismic impacts of big data, the necessity of data cataloging, and the transition towards cloud infrastructure, all within the context of AI’s growing importance.
Building a data-driven culture within organizations is no small feat, and Felix provides invaluable insights into this demanding process. Learn how shifting the role of data scientists to AI stewards and fostering collaboration between technical and business stakeholders can unlock AI's full potential. Felix shares real-world examples, such as Heineken’s success in creating a common language to facilitate digital transformation. We dive into best practices for cultivating a data-first mindset, stressing the importance of clear communication and shared understanding in aligning complex organizations to maximize AI impact.
Finally, we turn our attention to the concept of data product thinking, starting from business challenges and working backwards to define the necessary data and technology. Felix discusses the crucial role of data offices in ensuring governance, quality, and accessibility to maintain relevance. He also highlights the importance of aligning data initiatives with mission-critical priorities to drive strategic value. Wrapping up our conversation, Felix offers a glimpse into Calibra's unique customer-centric culture, shaped by its European roots and expanded to New York, underscoring their commitment to customer intimacy and a humble yet passionate team environment. Don't miss this episode if you're keen on understanding how to harness data governance for AI success!
More at https://linktr.ee/EvanKirstel
Speaker 1:
Hey everybody, fascinating discussion today on the role of data governance in this new era of AI. Felix, how are you? I'm doing really well. How are you, evan? I'm really well. Thanks so much for joining, really intrigued by your mission and vision at Colibra. Maybe start with introductions to yourself and the team at Colibra.
Speaker 2:
Absolutely so. I'm Felix van der Malen, co-founder, ceo of Calibra. Calibra, we're a software company focusing on data governance, data intelligence, basically doing more with trusted data. We help large organizations across any industry understand what data they have, how to trust it, how to use it effectively, how to make it easier to find the right data and, as you can imagine, as you hinted at, in the world of AI, data is more important than ever. Trusted data is more important than ever. There's only more data. There's more people that want to do something with that data. There's more use cases that require that data. There's more scrutiny on that data from a privacy, security perspective, and so doing data at scale in a trusted way is incredibly important, and we've been doing this for 16 years now. Started the company in 2008. So we've been at it for a while, but the data has only become more and more important, so still super excited of the work that we're doing.
Speaker 1:
Yeah, today's your day to shine and thrive, but talk about the role of data governance from when you started, maybe, through to the current era, and why is it so crucial for all users to access all data?
Speaker 2:
Absolutely, and it's only become more important. We started the company in 2008, actually a spinooff from the university of brussels. Uh, the four founders were doing academic research on semantic technology, which is now again and invoke. But, uh, how do you get people together to agree on what data means? And that's kind of our background. We were really passionate about that, but data wasn't as big as it is today. Uh, 2008. So we started 2008, as you can, 2008,. What makes you think of 2008?
Speaker 2:
Financial crisis and actually financial crisis helped us find an issue product market fit because all of those large financial institutions had to suddenly prove to the regulator that they were in control of their data. They understood what data they had, who was being used, how they were calculating their reports and their key metrics. So data suddenly became a business problem. And controlling data? So data suddenly became a business problem. And controlling data, data governance became a business problem. And that's where we initially started, where data governance became a big topic. There was a new role in organizations the chief data officer very much tasked with hey, help us control, help us comply to all of these new regulations that we are now being faced with, because the regulators are asking us prove to us that again you're in control of your data. So that was kind of step one between 2008, 2010, 2012, where that was kind of the key first phase. The other kind of big change that we've seen then is this trend towards self-service analytics. At the time it was Tableau and Qlik and now you have Power BI. Everybody wanted to use data to do their job. Everybody became an analyst. Data democratization, access to data for everyone, which was fantastic. But you can imagine what happens as it becomes chaos. Everybody has everything. So there's 50 different versions of the same report with 15 different numbers, and then the question becomes okay, why is my number right and your number wrong? And that became really problematic. So again, you needed governance and trust for analytics, for reporting.
Speaker 2:
The other big shift that we saw at the time was big data. Suddenly, everybody really. Well, actually, the more data we have, the more we can kind of create value from it. But again, as you can imagine, organizations started to capture more and more data. The problem wasn't do we have the right data? The problem is, how do we find the right data? The analogy I like to make is that your haystack, your data, your data lake has gotten a lot bigger. The needle you want to find that data set, that you need to do some analysis, build a model, is still as small, so you actually made it harder to find the right data. And so, again, that's where data cataloging became really important, or data marketplace how do we help everybody find the right data so they can actually do the job they want to do with that data? So that was another really important kind of shift in this data governance. From a role perspective, the chief data officer became the chief data and analytics officer because it wasn't just playing defense and regulatory, it was also playing offense and, okay, how do we help our organization become more data-driven? So that was a big shift.
Speaker 2:
And then over the last five years, of course, the move to the cloud data modernization. Every organization is modernizing their infrastructure, moving from on-prem legacy tooling to the cloud to deal with that volume, that complexity, that performance, that scale. And again, you can't move things to the cloud if you don't know what you have. Organizations are always very concerned that all the controls that they have on their on-prem environment, we have them on our cloud environment as well. So again, data governance became incredibly important to help in those data modernization trajectories and now kind of where we are today.
Speaker 2:
To no surprise to anybody, ai is the big driver and we've had data privacy in there as well, with GDPR, ccpa, again, sensitive data classifications. How are we using that data? How are we storing that data? And now in the world of AI, I think it's been clear now that data has probably the biggest impact on the quality of a model. What data are using to train your model and how is the quality of the data impacting the quality of the results? A lot of organizations are using existing commercial open source models and they're using their own proprietary enterprise data on top of those models. Again, how do you deal with privacy concerns? How do you deal with security concerns? How do you make sure that the quality you're using of that data is correct? How are you monitoring it? How are you assessing risks? How are you validating that? This is something that you actually want to do. So all of these problems again with AI is only kind of bigger and bigger and bigger, and data governance is only becoming more and more important.
Speaker 1:
Well, that's quite a lot to unpack. You know, fascinating overview. Talk, if you would, about the role of the chief data citizen, something you talk a lot about. I haven't heard that term of art yet, but elaborate on the roles, responsibilities for the chief data citizen and how it's different from you mentioned. Chief data officer, chief analytics officer, these kind of more traditional roles.
Speaker 2:
Yeah, and it's really. It goes back to kind of our mission and vision where we believe that everybody in an organization ultimately needs data to do their job. That's kind of the everybody has become a knowledge worker and we call knowledge work, we call it a data citizen. Right, everybody has become a data citizen and so as a citizen, you have rights and responsibilities and we really like that kind of balance. And as a data citizen, you have the right to easy be able to access high quality, trusted data and we should make it as easy as possible for anybody to do that effectively, because it's so important to do their job effectively. Right, and that's your right as a data citizen easy access to trusted data. But as a data citizen, you also have a responsibility. You have a responsibility to treat that data appropriately right. There are privacy concerns, there are confidentiality concerns, there are security concerns and that balance is really important.
Speaker 2:
But a lot of organizations, if you think about it, their biggest challenge with data is not having a bigger database, right, it's not having a faster database.
Speaker 2:
If they're not able to get value from data, it's not because they don't have the cloud right, it's because they still have to build that culture in their organization, that data literacy, that data-driven culture, to actually train people of how to use data, where to find data, what they can do with data, and so to kind of accentuate that fact, we believe that this role of a chief data citizen in practice that's typically the chief data, chief analytics officer but just to double down on this importance of building that data culture, building that data literacy being probably much more impactful in your ability to get value from data than building that data literacy being probably much more impactful in your ability to get the value from data than, again, having a bigger or faster database, and so that's why we really kind of doubled out on that role. We have our own chief data citizens. One of our co-founders, internity, who runs our data team, make sure that all of our equilibriums, as we call ourselves, are very data driven, and that's something that we kind of really advocate and encourage.
Speaker 1:
Oh, fantastic, and data science is changing fundamentally to a very different role today and in the future than we've seen in the past. Massive growth expected across this field. But how do you see this transition from you know a data steward role as a data scientist to kind of AI steward, and what needs to change within the organization or the culture?
Speaker 2:
Great, great question. And so, again, over the last few years we've had the rise of the data stewards right, to make sure that we can trust that data almost as the glue between the business stakeholders, knowledge worker and the technical personas, and you definitely need that kind of glue to kind of bring those teams together. And now, with AI, I think you're going to see a very similar needs. You have the very technical machine learning engineers and you have the business problems, the business use cases of what can we actually use AI for that drives business value. How do you bring those two together and do that in a way that is appropriate from a risk perspective, from a cost perspective, from a business value perspective, right. And so I think the data stewards are actually really well positioned, because so much about AI is really data. I think there's this saying that AI is really just a user interface on top of your data, which I like, and it's sometimes true. So data stewards have an enormous opportunity to jump in into that kind of AI steward role, if you will, to also facilitate and facilitate the collaboration between all these different stakeholders, which is absolutely critical to be able to put an AI use case all the way from kind of prototyping, discovery, all the way into production, because what we've found and what we've seen with all of our customers that we work with is that it's not an easy process.
Speaker 2:
Prototyping around AI.
Speaker 2:
Especially now with Genentech AI is easy and that's a big unlock where it's so easy to use ChatGPT and Cloud and all these other models to have a quick prototype, and it's really compelling GPT and cloud and all these other models to have a quick prototype and it's really compelling.
Speaker 2:
We're getting that from 80% quality to the 95% or 99% quality that it often requires to actually be in production is actually a big lift. And there's a lot of concerns around again security, privacy, risk, all these people, legal finance they all need to get involved to push an AI use case from again development into production, and typically a data scientist or a machine learning engineer is not the best person to do that. One, it's not typically their core strength. And two, you want them working on building great models, great use cases and training. You don't want them necessarily having to do all the collaboration with all these different stakeholders, and so that's where, again, there's an opportunity around AI governance. This is why it's not just around kind of risk control, but also how to accelerate the process of putting an AI use case in production and for an AI steward to help facilitate that, so kind of accelerating the ability to kind of innovate with AI in organizations.
Speaker 1:
Fantastic and you work with almost a thousand companies, customers, in this space. Talk about a data-driven culture. What are best practices? What have you seen that works or doesn't work, both internally and externally, when they're facing customers, partners, et cetera?
Speaker 2:
Absolutely. There's a couple of things. We work with a very large, again large Heineken, which I'm sure we all know and appreciate.
Speaker 1:
Very familiar.
Speaker 2:
Very familiar, great company to work with. They've been a long-term partner and they've really built their own kind of language, a shared language, because digital transformation is a key part of their strategy and either have a shared common language to be able to kind of drive a digital transformation like what do we mean by the consumer? What do we mean the customer, what does that really mean for us? How do we address that? You need to drive alignment and so building that data literacy, that data culture, often start by making sure you actually understand each other and you build that common language in a very large, complex, international, global organization. Right, so that that's a really important part. It seems simple, but it's really cornerstone. The second best practice that we see with our customers is that thinking, moving backwards and so, starting from the data and starting from the technology, start from the business problem or start from the business opportunity when do we have an opportunity to use data for? What is the use case? And then work backwards, backwards what data do we need for that use case? And that's kind of how you drive your data, that's how you, how you decide what to work on, and there's a trend recently that we're seeing a lot of uh traction, which uh, which we fully align to, is this data product thinking like how do you, how do you take product product management thinking and apply it to data? And product management is an opportunity. You build a business case, you do exploration, what is the solution, and you really build a complete product. A complete product is not just the software code that you write, it's usability, it's documentation, it's go-to-market, it's pricing, it's a whole package, right, and you should think of data in the same way.
Speaker 2:
The value of data is not just a table in a database. That's not the value. The value is, if I want to do customer-churned analysis, where do I find a customer-churned data product that I can use? Yes, there's a data component to it, but it's also. Is there documentation? What are the SLAs? What are the SLAs? What are the privacy concerns? How am I able to use that? Are there examples that I can easily use? Is there documentation that I can use? If I have a problem? Where do I go to? What is the data quality?
Speaker 2:
So all these concerns, like actually building a data product, is really important because it drives consumption right. It's really focused on the consumer of the data versus the producer or the technical stakeholder of that data. That's one reason why this is really important. And two, it helps you scope and focus on what's really important, and nowadays we live in a different economic environment.
Speaker 2:
We all feel the pressure to do more with less, and focusing on actually what matters and focusing on what drives a business impact is absolutely critical right, every team, and data teams as well, are under pressure to show like hey, how can we be more efficient, how can we do more with less? So making sure you focus on the things that are really important is critical, and that's where data product thinking is really important as well, because you start from the business opportunity, you start from a business problem and you work backwards. And sometimes when I see data teams fail is because they have too much of a technical thinking and they build something and they build a better mousetrap and they hope the people will come, versus starting from the problem or the opportunity and working backwards.
Speaker 1:
Oh well, said Speaking of which, you published something called the Data Office 2025 Vision. Maybe you could describe what that is exactly, some of the key elements and how you see it shaping or informing the future combination of playing defense and offense Croutony around security, privacy, compliance risk is not going to go away.
Speaker 2:
It's only getting more. So you really have to make sure you're doing the right things to kind of be able to use data well. But also, how do you drive value? And again, this data product thinking is really important there. How do you focus on the things that matter? How do you focus on not just the data, not just having again a bigger database, but how do you focus on making it easier for people to consume that data, trust that data as a really important component of there?
Speaker 2:
And then third is also AI. Again, we talk a lot about AI, a lot of experimentation with AI. Data is going to be absolutely critical. I think it's going to be the biggest constraint of actually getting AI use cases in production. And I think it's going to be the biggest constraint of actually getting AI use cases in production. And I think it's now on the data office to kind of step up. I would argue step up or, frankly, become irrelevant. Because if you don't step up and really help the organization be successful with AI again, by, one, controlling risks during the governance but two, making all the data easily available to be used in those models. I think your role is going to be, I think you've missed an opportunity, and so the data office of 2025 needs to play a really important role in this kind of new world of AI, both from an AI governance perspective, from an AI quality perspective and from a data product perspective governance perspective, from an AI quality perspective and from a data product perspective, Fantastic.
Speaker 1:
Any advice to aspiring data leaders on some best practice at the operational level or strategy side? You see so much across different industries and companies. What advice would you give?
Speaker 2:
It's a great question. I think that it's a bit of a cliche, but I would argue people process still super important. Change management, building that data culture is incredibly important. Technology is only going to bring you that far. Internal communication, internal marketing, change management again is really important.
Speaker 2:
Two linking what you do to something that is mission critical for the organization, like what is the strategy of the organization, what are the top three priorities of our company and how can I, as a data leader or a data team, support that right and again work backwards, whether it's through a data product approach, a data mesh approach, that's really really important. It's too easy to just again build foundational capabilities, but then the question, especially in this environment, becomes okay, what's the impact? Is this absolutely critical? What would happen if we don't do this? That's why it's so important to be able to tie yourself to a business-critical initiative, and there's no more business-critical initiative, I'm sure, with many organizations than AI today.
Speaker 2:
Right, everybody's rethinking okay, how is our business going to change through AI? Are we going to get disrupted? What are the opportunities to do things differently? So absolutely tie yourself to that opportunity, jump in. There's a lot of unknown, but it's through doing that you'll learn, and so makes also sure that you're not too much of a bottleneck right. You don't want to be the no person Again. Compliance is really important. Risk management is really important a big part of the job but also making sure that you enable people, you enable data scientists to get their models in production. That's how you're going to create a lot of goodwill and impact in the organization.
Speaker 1:
Fantastic insights. So you work across so many industries, from consumer goods, heineken, through to healthcare a personal passion of mine. You've helped I see recently UCLA Health with putting reliable, high-quality, clean data in the hands of healthcare data citizens. Firstly, that must be very satisfying to work in the healthcare space, but it's so important in this fragmented world of healthcare siloed world that must be a key area for you moving forward.
Speaker 2:
Healthcare pharma is a key area for us since the very beginning. One very data-driven, if you think about the R&D process and pharma Compliance privacy incredibly important, right. Patient information is incredibly sensitive. Again, how do you enable those researchers or people to do more with that data, to be more efficient, discover more drugs, do that faster, while also making sure that they're able to do that in a kind of compliant matter? So we work with a lot of very large pharma companies. We work with a lot of healthcare, healthcare systems, healthcare insurers. It's an incredibly important space. And now, again with AI, we have a few customers in that space that are really kind of on the forefront of AI.
Speaker 2:
It's been a while. We always think like AI is new. Of course AI is new, generative AI is maybe new, but they have done AI for a long time and they've done AI governance with Colibra already for a long time. Because, again, the way you do that is incredibly important. You want to get beyond the notices to risk you, and how do you do that? By putting in the right controls. And so, again in healthcare pharma, a tremendous opportunity to do that well.
Speaker 1:
Wonderful work and give us a peek behind the curtain at Colibra. What does the life of Chief Data Citizen look like? Or yourself, describe the culture, if you would, for the folks listening and watching scale leader in our category.
Speaker 2:
We care deeply about our mission and our vision. I think doing more of a trusted data, like I said, is only becoming more important. So I think we have a really important job to do with our customers. Culturally, I think we're very customer-centric. It's in the very beginning anecdotal story. We started in Belgium Very quickly. We expanded to New York.
Speaker 2:
At the time New York was not the tech hub it is today. It was very much San Francisco, silicon Valley, even Boston a bit more on the East Coast, but we followed our customers. We didn't follow our investors, otherwise we would have probably ended up in Silicon Valley, but we followed our customers at the time, large financial institutions which New York, obviously, is the headquarter of. And it's a very anecdotal example of how customer intimacy, customer championship is how we call it is a big part of our culture. The other thing you probably hear is I think we have an interesting combination of our European roots with the ambition and the drive of where we are now in the US. I'm in New York since the last 10 years, but I think it's an interesting combination. We try to make it the best of both worlds and combining that it creates, I think, a unique culture, which I think is really, really attractive by having very passionate people that care deeply about what we do, but also with a level of humility that I think is important as well.
Speaker 1:
Wonderful. Well, as someone who lived in Belgium a couple of years in my 20s, I worked in Brussels and lived in Leuven. I hope you're going to get out to the cafes and the restaurants in Belgium this summer. As you know better than I it's a wonderful time of the year to be there. So thanks so much for joining and appreciate you sharing the message and the insights and the education, absolutely, and thanks for having me. Thanks so much, thanks everyone. Take care, take care.
