All right. Hello, everybody. Welcome Welcome to our webcast on the importance of data governance and financial services. This is going to be fairly, I want to say, casual conversation.

We’re not going to use a whole lot of PowerPoint. In fact, we’re only going to use one slide just to introduce the topics. Most of this is going to be just a panel conversation between myself and the three folks that I’m just about to introduce here. Our general topic is importance of data governance and master data management for financial services.

So we’ll get into some things that are specific to financial services and a number of things that are more general around the whole topic of governance and master data management. We have three I would say all three are very much experts on the panel here today. And I must say, ever since we put this presentation together and we’ve been doing a little bit of prep around it, I’ve been very excited to have this conversation, because there is an awful lot of experience on the call here today. So let me introduce the folks here and have them say a little bit about themselves.

We’ll start with Patrick Terry. Patrick is the chief data officer of the Bank of Oklahoma and he’s on his fourth, currently working on his fourth master data management implementation where he’s been doing this for a number of years. Patrick, can you tell us a little bit about yourself and some of your MDM credentials and I guess financial services credentials also? Yeah, Thanks for having me.

I have been in the data space for almost 30 years I’ve worked at large banks and smaller banks from Wells Fargo to first horizon to Heartland financial and now B O’Keefe. I have done, as you said, for different or in my fourth implementation of MDM, I’ve done two customers and one vendor and now a second vendor. So I have a good sort of cross platform, different approaches and different technologies and lots of learnings in this space over the last 20, 25 years. Always been always led off with customer domain in all of these cases.

Excellent and I’ll point out that as chief data officer, Patrick has a very specific interest and focus around the role of the chief data officer. I think in the middle of writing your doctorate dissertation around the role of the chief data officer. Right, Patrick. I am.

And trying to help the industry with understand the value of the data officer as chief data officer and how we interact with other executives. And some of the challenges. We have been pretty new, new c-level role in the industry. So we’ll dive into a little bit of that conversation as we get into the conversation.

I think it’s an excellent perspective for anyone here our audience to have. Next panelist Pedro Franco, who is the director of data and AI at New desk and geodesic, is a proxy partner. So they specialize in data strategies. Actually, I’m going to let Franco talk about what new desk specializes in, but new desk and Franco have done prophecy implementations as well as governance and with purview and other platforms.

And Pedro, why don’t you introduce yourself and what Udacity does for living in what your specific role within that is? Absolutely Thank you for the introduction. Martin let’s start with your basic. We help our customers accelerating their digital business transformation journey.

I lead a data and I practice within the data I practice. We concentrate heavily on data science and engineering and architecture. And a year ago when I joined, I realized that we are about to hit this massive need for data governance. As the data stakes keeps growing, as we keep getting more and more information is it’s important to start putting a priority on data governance.

That’s one of the keys that I’ve been concentrating on my experience. So as a data professional for 15 years, I’ve worn many hats. I work as a Etl database and business intelligence developer, then moving to the database administration side of the house. So understanding the infrastructure, security and networking component of the profession.

And then I move into data architecture and it is 8 years ago that I actually had the opportunity to manage an organization’s entire data state, and through that I learned the importance of master data management. Since then, I’ve been concentrating on producing architecture for our clients and championing data governance, which I consider like two of the main parts of the foundation are metadata management as well as master data management. Excellent. Thank you, Pedro.

Our third panelist here today is Malcolm Hall. Who’s at prophecy. And, Malcolm, I’m just going to let you introduce yourself. Great thanks, Martin.

I’m Malcolm Hawker. I’m the head of data strategy with prophecy. I’ve been in my role now for just over two months. Prior to that, I was with an analyst firm called Gartner and the co-author of the last three MDM Magic Quadrant documents and a number of other documents related to best practices around MDM data governance.

In my role as an analyst, I was given the privilege of talking with literally hundreds and hundreds of CIOs and cdos, including many in the financial services space, about the role of MDM and about the role of governance in today’s modern organization in today’s increasingly focused world of digital transformation. I’ve had so many interesting conversations about what’s working and about more importantly, what’s not working. Prior to my tenure with Gartner, I was with a small data provider called dun and Bradstreet. Prior to that, I was actually running an IT organization for $2 billion publicly traded company where I stood up MDM and a governance organization.

And in many ways learned a lot of the basic blocking and tackling of MDM and governance the hard way through a number of actually fairly significant challenges. And I’ll say it, failures in the past are related to MDM and governance. So over the last couple of decades I’ve been firmly rooted in the data and analytics space and in the data management space. And I’m so excited to be here today and to have a conversation with, with a couple of learned peers and share information about, again, what we know works and what we know does not work in this space.

Excellent. Thank you. And I’ll point out that Malcolm, now at prophecy is really head of data strategy, but also evangelizing a lot of Indeed just how to use MDM, when should it be used. And just evangelizing the whole idea of MDM starting a podcast, very active on LinkedIn and definitely a resource for all of us as we go forward.

So I would encourage any one of you to reach out to Malcolm on LinkedIn if you’re so inclined. So I’m going to share just one slide here on what we’re going to talk about today. As I said, this is very much intended to be a discussion, not a presentation. So this is the only PowerPoint slide.

And I put it up just for the audience. And quite a few of you have joined since we started the conversation at the top of the hour here. These are the discussion topics we’re going to go through. Digital transformation, how it’s impacting financial services and how data should be managed.

We’ll talk about because it is I think we all recognize very much a top of mind idea. Gartner says digital transformation is a top 10 issue for CEOs and PriceWaterhouse. Recently in a survey said 76% of CEOs consider digital transformation the number one priority. So we’re going to talk about how digital transformation is present in financial services.

What are the drivers, what are the particular things that are relevant when it comes to MDM and governance? We’ll also talk a little bit about technology trends, which trends are driving financial services and which are not things like cloud machine learning and artificial intelligence and again, governance and master data management. And then a practical hints and tips from what is clearly a very, very experienced panel here on how to make it happen. What does it take to put this into practice?

What are some of the things that you as practitioners in the audience should be thinking about? So we’re going to go through these conversations here, and I would encourage anyone in the audience to drop a question and to chat if there’s anything you would like us to expound upon or get into a little bit more detail. And we will do our best to cover anything that you drop in there. OK with that, let me share that screen and get started on some conversations.

So the whole digital transformation topic. We’ll start with Patrick, who’s very much, I think, thinking hard about how to navigate bank of Oklahoma through digital transformation. What would be some of the things that are top of mind with you, Patrick, in terms of what’s driving the business with regard to using data to accomplish business outcomes? Yeah, I mean, certainly with any digital transformation project, generally, there’s some type of system that’s being implemented.

But behind those systems are connections to other systems. Or for any system you implement, you have to connect to multiple places. And that’s where our world comes in. When you’re needing to master data or bringing different data systems together, cleanse data, good clean, quality data.

Those are. Things that often get a little bit overlooked when projects start. They’re excited about starting an initiative and they start standing up the platform and then don’t always think about all the data integrations and the normalization or granularity of data that’s going to be needed. That’s where we have found some challenges and opportunities to, to try to think about earlier on and projects or have to deal with challenges in data where master data management can be a good solution for it.

Any particular experiences in terms of where data has been used to drive some kind of change in business or increased efficiency that come to mind when we talk about this. Yeah, I like the customer data story. Pretty much every MDM implementation I’ve ever done has started with customer financial services companies. They create a relationship through an establishment of an account and it’s basically an account to customer relationship, and that’s pretty backwards on the way that we want to sell customers or service customers.

We want to service the customer, not the account. But yet systems are designed to start a new checking account or start a new loan process. I mean, at the heart of that is customer data. And if you have more than one relationship with a financial services company, chances are it’s going to be in two different systems.

And needing to understand that customer and roll that up between two systems seems to be fundamental and table stakes. I’ve learned not so much, always an initial focus. And so that’s where master data management and customer kind of come together, is to sort of dupe that customer or to provide that holistic view of the customer no matter what their various relationships are. So you’ve always started from your perspective in all four of your implementations with customer and using it to drive, using MDM, to consolidate data, to drive the customer experience and cross-sell upsell and effectively manage the customer’s relationship with the financial organization.

Yeah, the last two implementations that I have done have been where they didn’t understand the customer as a whole. And some maybe older implementations of India might have been centered around just analytics and rollout. But what I have found, where I am today and where I just came from, the last two implementations, it was operational. It needed to be near real time that having these Roll Ups and these combined relationships and then available to a system in very near real time CRM solutions that are customer contact centers, where you’re engaging a customer either in a sales process or a service or support, you need real time access to all of the relationships that they have with your organization.

In India, an operational is at the core of that because where else do you bring all your systems together in one place to service yet another system that’s trying to represent all of those relationships? And that’s where MDM and operational in them is I like to call it really, really plays in. Excellent yeah, that makes a ton of sense. Let’s broaden it out a little bit and ask Pedro to talk a little about the same kind of thing.

Maybe some experience you’ve had, Pedro, outside of financial services or any place where there was bringing together multiple data to accomplish a business objective. Just to give everyone a perspective on how MDM can be applied here. What what are some of your experiences here? I agree with everything that have been said so far.

Master data management actually enables data interoperability as well as data quality. So the data interoperability piece is getting the entire organization to speak one language. Whether you go with data fabric or data mesh or any other type of architecture currently in the market without a master data management strategy, it’s really hard to reconcile all those data assets in the organization to get everybody to speak the same language. I had the opportunity to implement master data management in Pride and logistics and transportation or travel industry.

The way that it helped was to improve customer experience. So having one United front on every single system that the organization had when communicated with the customer, and this could be facilities, this could be locations not having this jolted, one of the very first problems that I faced was customers would come, they will receive emails, they will receive messages in text, via text. They will go to a website. They will go in different aspects of the organization.

Communication strategy and the name of the location where they were supposed to go was different. So the experience was already tarnished just because they couldn’t understand is it one thing or the other? So implementing master data management in that place, make sure that it was a consistent experience no matter what channel of communication. We will use.

In the other side, it was slightly different. It was being able to have every aspect of the organization come to one place to get the same data points. So when they will go to their system, if they needed to reconcile, they would use the master key as opposed to have some sort of mapping exercise to make sure that this ID means the other ID and the other system and have to go through data cleansing while they’re trying to do their analytics. So from an analytics perspective, you have all these one version of the true unified, all these data points that you can have about your customers or any other asset in the organization.

And from an operational perspective, which is something that I saw recently, was reducing the processing costs. So the use case was to have a transactional systems and the millions of messages they check the MDM tables for making sure that those messages were correct. Unfortunately, that master data wasn’t updated in real time. Unfortunately, that master data was updated once a month and what it cost was insane costs and time.

Just reprocessing those messages that didn’t pass validation one month later. So extremely reactive, extremely time consuming and cost consuming. So those are my, my experience implementing MDM and how to solve various problems in the organization. Excellent Thanks.

So, Malcolm, let me ask you and we’ll refocus back into financial services. Obviously, Patrick’s focus is around customer and that’s something that we see very often. But I think you’ve probably seen and heard of other use cases for MDM and governance within financial services and you kind of widen the aperture a little bit within financial services and talk a little bit about what you’ve seen and what other things may come up. Yeah, yeah, absolutely.

And expand on a couple of things that I’ve heard that I also completely and totally agree with as well. So from the financial services perspective, yes, kind of customer is most certainly the benchmark that most are aiming for. And what we’re seeing, though, is increasing increasing awareness that there are multiple views within the organization that are all potentially correct. So Patrick noted this notion of customer versus account.

And what we’re seeing is increasingly marketers want both. Right and maybe you’ve got commercial customers as well. Right? and maybe those tax IDs is the right view, but maybe more of a corporate headquarters view is the right view.

So we’re starting to touch on areas of hierarchy management here as well. So marketers increasingly want to have the ability to look any way they want at a customer in multiple ways at a customer. And all of those views could be accurate, given the needs of a specific campaign or specific product or a specific outcome that you’re aiming for from an MDM program. So yes, all those things are absolutely true and it’s no longer kind of a world of one view must rule them all right.

In the banking space, particularly commercial space, it was tax ID, that was it. This is the ID to rule them all. And increasingly we want we’re seeing organizations want the flexibility of looking multiple ways. Right and having a system that can support that and having a hierarchy that can support that was also mentioned the importance of operational MDM couldn’t agree more that’s where most of the value will be realized is creating that data in an MDM hub and then propagating that data down into a CRM system or some sort of customer support system or an ERP system.

Absolutely agree. But for smaller banks that are coming late to the MDM area and many are many still are always having a lot of conversation previously at Gartner with smaller banks that are just now trying to figure out how to implement MDM. And for a lot of these smaller players may be more of an analytical approach is a good first start right where all you’re trying to do as a step one. Yes step two or step three.

The goal here will be to get to that operational MDM bidirectional flow of data between source systems and an MDM hub. But maybe as a starting point simply just, you know, creating aa3 60 of something, 360 of a customer or a single view of a customer vis a vis some sort of report could be very, very powerful, right? Even simple ease cases like understanding what’s the crossover potentially between our consumer space and our commercial world. There’s a very, very long tail there that could be very lucrative for marketers to understand where there are potentially overlaps there.

So certainly the customer experience is really kind of driving a lot of the digital transformation efforts within financial. Services but, Martin, to your question about other use cases, you know, some of the old favorites are always there. Right and staying abreast of KYC and staying abreast of compliance and any sort of regulatory driven pressures here are obviously still driving MDM programs. They lead banks to be one of the early adopters of MDM, and they lead banks to be, frankly, ahead of the curve when it comes to a data governance perspective.

So those use cases are still there because our regulatory environment, as everybody on this call knows, is a dynamic thing. It’s always changing. So those use cases are certainly there. What I was seeing more and more Gartner is, interestingly enough, is more and more of a focus within banks, within what we would call the product domain within MDM, which is I think is a natural consequence of more and more and more and more financial services being pushed online.

Right and having continuity between the product experience that you would have within a brick and mortar location or some sort of online location. But where banks are seeing now actually the benefit of product centric MDM and all of their various offerings. So there’s a couple other areas of focus that I was certainly seeing from financial services organizations, but without a doubt, the biggest driver here most certainly is the customer experience. Excellent all right.

So there’s been a few questions coming in about governance. So let’s pivot to that and start talking a little bit about technology and how it all fits in. So, Patrick, let me start with you. Can you give a perspective on how you feel about governance in financial services and in your current situation?

What does that really mean? As a practical matter? What is governance all about? How many systems are you trying to govern across and how do you go about something like that?

Using technology and also using the organization. And all of the different factors that would be required to make it work. Yeah, glad to respond to that and I’ve read both of those questions. Data governance is a tool.

Data governance is a process. I mean, it is a combination. When you when you think about data governance and financial services, it’s how you comply with all of the rules and privacy and regulations, how you help assure clean data for those, how you ensure privacy of data. I mean, data governance encompasses a lot in India is one part of that ecosystem.

Can you implement without it? People have for years. Should you implement without it? It depends on your situation and the skill of the company.

So that’s it that the answer to that? It depends. Master data management has its I mean, where it really brings value is how it will bring domains of data together and how you can use the act of MDM to do data quality. For example, in my last this one and the last one, it was our first time that we’ve done address cleansing.

For example, just taking our entire address database. And if you’re trying to match merge on something related to address and you don’t have clean addresses, so take it an opportunity to apply good data quality and data governance against stuff that is important is sure and cleanse and clean addresses taking those opportunities. So that’s how it plays together. But you definitely data governance is it’s much more of a framework where technology is just one spoken that will, between the processes and the people and data ownership and stewardship.

There’s a lot of parts and pieces to an effective data governance program where MDM does play a role, but it’s just real world. So Patrick so there’s an interplay between people, process, and technology. Do you have a perspective on is there always one right place to start or do you start with all three simultaneously? Or how important is the toolset that you pick?

I mean, how do you how do you think about getting started on that? Yeah and I’ll use this as an opportunity to say don’t let MDM technology be a solution looking for a problem. Right? do it the other way around.

So to answer your question, I would say start with the processes and the people. Understand what the needs are, what the business requirements are, prioritizing the data governance demand, just like any other demand. What is the most important thing that we need to govern and where do you start with that? What is the current processes?

What processes need to improve? What people are involved? A big part of it is the workflow, right? It’s one thing to master data, but another big benefit a lot of people don’t think about with master data management.

And part of the reason we’re doing it at boc, we had a way to do in the past and that’s part of MDM. But part of what we’re looking to achieve with our current implementation is the change control, the tracking of change, the approval of changes of, of particular critical data. The elements that you’re governing if you’re going to change an officer. Has that been approved?

Have you have you checked off all of the boxes in terms of the process behind that? So so you start with the business processes. And you start with what that demand is and then look for what technology within say in this conversation in them can support you on. Super so, Pedro, a similar question to you then.

I know you’ve got experience and you’re standing up the governance practice within Odessa. What’s your thoughts on how to implement governance, where to get started, tools, technology versus organizational imperatives? Well, going back to that question mark, is it possible to have a DG program without implementing mdm? In my opinion, no.

You will not have a data governance program, a complete data governance program without having MDM. If you go by BMI international, which is the International non-profit organization that rules data governance or data management datum reference data master data is one of the core components. In my experience, master data management enables to other pieces of data governance, which is interoperability as well as data quality. Without having a solid master data management strategy in place, without having a solid and metadata management strategy in place.

It’s a little bit difficult to really manage data in an organization, a government properly. In addition to that, master data management also enables security and enable security by making sure that the systems that require certain core components of the organization and getting the best version of it so you don’t have master data or pieces of master, they are laying in different systems where people can misuse it. So there are many components that master data management enables in the data governance space that make the whole process work. I think it’s one of the core pieces.

Otherwise, your data governance strategy will stand in one leg or two legs. It wouldn’t be like the whole process in terms of how to go about it. When it comes to technology and when it comes to just the tooling aspect of it. I see the Union between azure, Microsoft Purview and prophecy master data management.

For someone that wants to implement master data metadata management in Azure as like the perfect fit the way the two tools work together. It’s incredible. From from a metadata management perspective, Microsoft Purview provides a lot of power. Of course they do much more than just better data management.

And then you have proxy MDM integrating extremely well with SQL Server and many other SQL Microsoft products. The way that I see how to go about a data governance strategy, I personally will start with metadata management. Really understand what you have in your organization. The answer or the outcome of metadata management is going to understand what is master data management?

I have countless encounters with companies where I ask, what is your master data? And they just give me a blank stare. They do not know what master data means. They don’t know what it is.

They said, well, we have this information, it’s this master or reference. And then I have to go in the whole explanation of what is reference data, what is master data? So having your metadata management in place and how in a well defined it gives you the starting point of identifying those entities in the organization that are master data. Once you enable your master data management strategy, then you have your success rate on all your analytical workloads is significantly greater.

Furthermore, you enable the entire organization to speak the same language. So how will I go about it? I start with metadata management going to master data management and then go into the analytics, the transactional loads and the big data loads. Excellent Malcolm, can you I saw you were nodding there.

So I think we’re all broadly in agreement that MDM and governance go hand in hand and there neither one of them really is all about the technology. There’s people process technology, but can you talk a little bit about what you’ve seen and how these things play together and really how to get started in building that up in your organization? Yeah thanks, Martin. I think confusion arises when we acknowledge that MDM is both a noun and a verb.

It’s a discipline. It’s a way of managing data, and it’s also technology. So to Pedro’s point, if you’re serious about data governance, you cannot do data governance without MDM as a discipline, because I don’t know anybody that doesn’t want to share data. Right like the MDM, if you were to distill things down to one or two pithy statements about what it really enables, it enables you to share the data.

And Petri is this wonderful metaphor of a common language, right? That’s what MDM is very as a discipline is very, very good at. So if you have a goal of data governance or if you have a goal of data sharing, then yes, absolutely. MDM is a critical component of that as it enables the other things that Peter mentioned as well, which is some aspects of data governance and uniqueness and consistency and trustworthiness and.

Also enables them some security and access as well. Are key components of the data governance framework. Now, we could have an interesting conversation about do you need do you as an organization need MDM as a technology? Because what all technologies are very, very good at is scaling, right?

Adding layers of automation, configurability, right? At a certain point, for most companies, the answer becomes Yes. We probably really do need MDM as a technology to enable the scale to go beyond a reliance on Excel spreadsheets or to go beyond a reliance on point to point integrations that are scalable and highly brittle. So for a lot of organizations, the answer here is yes, we probably do need MDM technology, but it’s important to draw a distinction there between MDM as a discipline and what role that plays in enabling a governance program and MDM as a technology.

In terms of what do you do to stand up a governance program or how do you get started with governance? Things have already been touched on by both Patrick and Pedro, which were which were which were bang on. So what Patrick was saying was, you know, be very, very outcome focused. Right have an outcome in mind, have a use case in mind that you’re trying to support.

Right what I see so often and what I saw when I was a Gartner analyst over and over and over again was companies would kind of take this bottoms up approach of governance, right? Where they would start by cataloging all of their data, inventorying all their data, profiling all their data lineages, all their data. And that process in and of itself can take months and months and months, particularly for a very, very large bank, and particularly if you don’t have a good understanding of where all of your data is or how it’s defined. So that approach can be time consuming.

And if you’ve got consultants engaged and 90% of the time when MDM software is being deployed, they are engaged. And we love consultants. But that can be a costly, a costly undertaking. So instead of the bottoms up, can you take more of a top down?

Right and this is mentioned by Patrick, which was, you know, again, focused on some sort of an outcome, focus on a limited set of use cases. Don’t try to be all things to all people. Master data is a great place to start from a governance program perspective because it acknowledges that not all data is created equally. There are certain data domains and certain processes that use those domains that are far more valuable and will drive more impact for the organization than just some random field of metadata sitting in a CRM somewhere.

So start from start from the top, start from some understanding of what the outcome is. Use that to establish some guardrails for what’s in scope and what’s out of scope on an initial phase governance program and then work your way down. Once you have your scope understood and your outcomes understood, then you can start the process of profiling and lineage and cataloging and clustering all of that data because you can be more focused and you will know that what you’re doing here is in service of one specific outcome instead of just kind of randomly chasing all data across the organization.

So that’s a great way to start governance, is to be very, very outcome driven, quantify what those outcomes are and use MDM as a great starting point again to some sort of Governance Initiative if you’re just starting from scratch. Excellent so I have some other technology questions that I want to ask, but since we’re on the topic of governance, I’m going to stay here a little bit longer. And Patrick, I know you’ve thought a lot about this in all of your implementations. What is the organizational structure or organizational buy in that’s required in order to make governance work?

Do you have how do you get the stakeholders all in the same page? What kind of meetings and agreements do you need to have at an organizational level that enables this stuff to the technology to come together and actually be successful? Yeah so I’ll address the how to part with data governance. I mean, there’s lots of good frameworks that Dana Denmark, I mean and there’s several out there.

The pyramid, the typical pyramid where you need that executive buy in, they need to understand what you’re doing and why you’re doing it, most importantly, what value it’s going to bring. And it starts there. Again, I agree with what you said a second ago. If you try to inventory everything and go bottom up, by the time you spend a bunch of money, nobody knows what you’ve really spent it on or what value you’re getting from it.

But if you take the other way around and data governance is formed from the leadership and you have every level all the way down to data storage that and people that are using the data or managing the data every day, and you create those effective layers and stewardship ops and different teams or groups. And I mean again, there’s lots of references for this but different frameworks but, but having the right people and the right conversations and not having everybody all at the same conversation, you can’t say. In the empty CEO. He won’t know or care what that even means or why.

But he might be writing and making lots of investments in it. But the conversation that the executive level needs to be about what problem you’re solving, what regulations you’re ensuring that you’re complying with, and then keeping the rest of the conversation is more technical at a technical level group. And so it’s really formulating those groups to have the right conversations at the right level and then making sure that it represents the entire horizontal across the whole company so that when things roll up and boil up to the executive level, they hear a consistent message from their verticals.

And that’s really how you do that and how you do it successfully. I like to say, you can’t speak to BLT. And that’s been an interesting lesson. As a CTO have come up through data, I manage well, I manage down very, very well.

And I’m still learning to manage up, to not be too technical or speak technical or speak about solutions. When I’m solving problems, I have to manage my message. Well and that’s I’m still in school on that one as a professional to be as effective at communicating data governance concepts and implementation to executives versus effectively managing the teams that are actually getting it done. Excellent we’ll talk a little bit more how to sell this to management when we get to our how to make it happen and put it into practice section.

Sticking on the technology then for now, impact of cloud technologies, adoption of cloud, is that something that’s necessary or a case by case or what’s your take on that? Patrick yeah, it’s absolutely necessary, but it needs to be evaluated on each use case, right? What is the business value? It goes right back to that cloud is not the answer to everything.

And staying on prem is clearly not going to be the answer to everything either. So it’s understanding what the value of cloud native technology and cloud technology can do for you and where on prem or even legacy dimensional type models are still going to be around for a long, long time. And maybe at some point they all end up as we live and breathe and grow and evolve, we have to realize that we’re going to be in a hybrid world for a really long time, and it’s knowing what questions you’re answering and whether it’s cloud technology or more on prem or what people like to call legacy technology.

I wouldn’t call it legacy, just not cloud or on prem type technology. There’s a place for both. And so really understanding and separating. The what problem or what question are you trying to answer?

Because especially with cloud, we’re learning that it’s need to separate compute from storage and the tools that you want to leverage. Some are cloud native and some work more natively better in a cloud environment. And so you see hybrid. We’re seeing on prem product cloud where we’re leveraging the power of cloud native technology on prem.

So it’s very complex. And so you really need to just break it into parts and pieces. And again, it goes back to knowing what problem you’re trying to solve. And what value it’s going to bring.

And then applying the best solution at the top. Excellent there are two verticals that we deal with that seem to be very sensitive to security, their health care and financial services. Is security a factor for you when you’re thinking about deployment method? It’s what prevented me for doing it for years.

A couple of jobs ago. The reason we didn’t and what held back a big data platform was the risk of security. I mean, when you think of the financial services, especially when you’re dealing with have everybody, you have all the information you could ever imagine from socials to account numbers to everything that can have identity stolen. There is no more industry more sensitive than financial services.

So to bring that data out of our firewalls into an Azure or AWS or any of the public cloud environment scares everyone. So that’s absolutely central to how you and when and if and then how you do it be on cloud and exist and leverage those cloud technologies. And I have found it all over the place in terms of comfort, level and expertise around that topic. OK do you think do you feel like the technology has turned the corner to the point where it’s possible to have trust in the security access controls that provides you?

We’re doing it today and comfortable with it. I’ve come into an environment where that was figured out before we went on cloud versus getting on cloud and then trying to figure out how to secure it. The partnership here was to partner with companies that specialize in securing data on cloud. And then we built a scalable cloud data strategy that at the foundation had already had a high comfort level on cloud in place.

So as long as we manage against that framework, there’s really no limit to our use cases of what we can put on cloud based on the security question that’s covered. And that wasn’t the case in a couple of other places I’ve worked with in the past. Excellent so for I guess, the same question, how are you seeing cloud as being a necessary enabler or an optional enabler of governance and mdm? Pedro what do you see most of your customers leaning towards, I guess, is probably a better way of asking the question.

Well, neovasc is a cloud partner, so most of our work is in Azure. I personally had to do the same journey when I was in corporate and working, managing a data state. There was a lot of security concerns, a lot of compliance concerns going to the cloud. I’ve spent many years of my career actually recruiting new technologies, technologies that come out of college.

So for me, one of the biggest drivers for adopting cloud technologies, at some point, you’re going to run out of talent. At some point, the guys that used to know how to manage on prem services are going to be retired, and at that point, you’re going to have to look at the New talent pool and the new talent pool for what I see every day. They want to work on cloud technologies. They always ask about Azure GCP, they always ask about various new technologies.

And when you ask them basic questions about what I used to do when I manage databases on prem blank stare. So if there is one driver that I see is how fast technology move, how these new generations are hungry for that new technology. So we’re implementing master data management unless absolutely you absolutely have to do it on prem. Also just going cloud now we have a lot of compliance and security built in on these cloud service providers.

In addition to that, you’re reducing all these administrative overhead that you can move to your data governance strategy. And now that everybody is moving to cloud data, governance becomes a business imperative because if you’re going to move to cloud, you’re going to have to deal with shadow ID, you’re going to have to deal with risk of people misusing data. And if you don’t have that strategy, well, enable before or at least a pretty good idea of it before you move to the cloud, you’re going to encounter a lot of trouble. OK, great.

Let me ask a different question for you, Malcolm. Use of artificial intelligence and machine learning in the world of MDM governance and what you can do with those things once you’ve done MDM and governance. Well, great question. And this is relevant to the cloud conversation as well.

Right so for a long time, organizations had a hard time justifying the Roi of a cloud migration because from a purely operations perspective, from an ops perspective, it wasn’t there. Right? the Roi just wasn’t there. So we can pay our bare metal or somebody else’s bare metal.

And the costs here are largely a hair split. But that world has changed. So there are ways increasingly there, and it’s increasingly there in the form of the insights that can be derived inferred from extremely large data sets using the technology that you just mentioned, whether that’s AI, ml, maybe it’s graph, maybe it’s something else. But what we are seeing now is marketers getting very excited about the insights that can be gleaned from extremely large data sets that would be poorly managed anywhere but the cloud, right?

So the cloud allows for scalability and elasticity in compute and storage, and we can store more and more and more data. And it’s not just storing data for data hoarding sake, it’s actually storing data to understand things like maybe a household relationship that you never understood before. You were able to apply some graph analytics into an extremely large data set, and you were able to see that two of your banking clients that you thought had no relationship whatsoever actually have a family relationship where some sort of household relationship that could be incredibly valuable, right?

So these very, very large data sets are now being mined in ways that have never been really that efficient before. Yes, we can talk about Hadoop and big data in the past, but what we’re seeing now is the application of graph into MDM use cases where we’re understanding the overlap from a relationship perspective of things that weren’t necessarily that compelling from an MDM perspective or may not even have been considered master data in the past, where now maybe they are being considered to be governed as master data because they are relevant from a marketing use case perspective. So the kind

of playing field is completely changed here, the use of these technologies, AI and ML again, whether you’re talking about fraud or there are so many risk mitigation use cases here as well that are rather compelling for understanding things like beneficial ownership and other areas of KYC that were not possible before, before having the data in the cloud and before having some of these advanced technologies. Yes, we can talk about how AI and ML are allowing automation in things like data stewardship, how they are allowing for added layers of scalability in the data management processes, how AI is allowing things like automation of data modeling and maybe even automation of creation of hierarchies, whether those are customer hierarchies and product hierarchies, the things are all very, very real.

But a lot of the needle here is being moved by some of those marketing use cases that I was talking about, because we’re able to derive new insights that are extremely compelling from a marketer perspective. Yeah kyc, fraud. Those others are also compelling certainly as well. And now, Patrick, I’m wondering your take on a.I.m. I know that you’re part of your justification for doing what you’re doing is that you can then apply those technologies.

I know that some of what you’re doing may be confidential, but what’s your what’s your general take on, on the use of those technologies in conjunction with master data management? Yeah, I think he said it very well. I mean, there’s a lot of things that we can do in our on prem environment, but most models need more data in order to make the model more accurate. And so, you know, when you start considering having two years’ worth of data and more of an active type space for reporting and dashboard and things like that, but for Melanesia to really have its teeth in real value, even if you’re looking at a very current data set to train the model, you may need 10 years worth of data.

And that’s where you really need cloud data or need a cloud platform to be able to go that far back. And then, you know, and the way that cloud technology has evolved now, you can see in most any data structure, whether it’s like comparing data in a more conventional model, it gets years and years of graph or other types of data space where you can train the models to be more effective and where we’re doing it heavily. You talked about marketing. It’s knowing your customer and knowing your relationships.

No one being able to mine your data for our next best offer. I mean, that’s not secret sauce but how people do, you know, what things you do might be a little bit but everybody is trying to get next best offer and we really can’t do that with on prem you could do some compares, but for ML and AI in the conversation for helping our sales team know what leads to go after. We have a team that that’s all they do is help our sales team know which leads to Chase and some of those our internal customers and some of that’s external data that we bring in and learn from. And in both of those cases, you need you need enough space that an on prem SQL Server is just not going to give you that.

All right. Excellent new topic for Pedro to kick off with since you mentioned earlier, I did a mesh data fabric. The question is around deployment architectures and how some of this data can be stored and worked on. Do you have any thoughts, Pedro, on what you’re seeing across your customer base?

I see that people are still slow to adopt data mesh. I think Pedro just froze. Pedro Pedro just froze. Malcolm, same question.

Oh, no, you’re back. Are you back? Pedro, say something. Yes, I’m here. Can you hear me?

I can hear you now, please. We heard your first couple of words, and then. Then you froze up. So please start over.

Going back to the question, it was mentioned Republic. My my point of view on that, what I see, what our clients are, clients are still adopting centralized data architecture. They’re still adopting data warehousing, they’re still going into the Data Lakehouse and having everything in one place. This new paradigm shift where we are starting to enable various parts of the organization to create their own data products is very progressive, but in my opinion, the only way that something like that can work is if you have a Federated data mesh where master data management sits at the top governing everyone.

Otherwise, data interoperability, which is one of the key components of the data mesh, will be broken. I am completely sure that over time, as we continue to democratize data and we continue to have this perception of data as a product or data service, data mesh will become the normal or the standard, but they will not be successful without a strong MDM strategy in place. That makes perfect sense to me, Malcolm. You have anything else to add to that?

I do. I’m a little bit of a data mesh contrarian. I’m far more bullish on the data fabric, and I can talk about why. But I would say easily that 80% to 90% of the conversations that I had when I was an MDM analyst were people trying to undo the impromptu mesh that they had created.

So this notion of domain ownership of data leads to silos. Right and if you need to understand how many customers you have at a corporate level or what your SKU inventory level is or what your supplier relationships look like at an enterprise wide level, you need some form of MDM to do that. Right and then and so. So yes, I understand the data mesh.

I’ve seen a lot of impromptu data meshes. I’ve seen a lot of companies now say, oh, well, this is what we have. We have a data mesh, we have domain centric ownership of our products. And we have a highly Federated do whatever you want from a domain perspective.

But a lot of those companies are struggling to try to get MDM as a way to resolve some of the shortcomings of that Federated approach without having any sort of enterprise wide view of that data. Now so obviously, unless you are a completely decentralized holding company that allows each of your business units or each of your departments to operate in a completely autonomous way, and that’s fine for you. There are some companies that operate that way. Most do not.

You’re going to need MDM as a foundational component of a data mesh. Now, data fabric I’m far more bullish on. I think that there’s some very, very interesting things that could be coming down the road in another 5 to seven years from the perspective of the use of active metadata to start fueling automation and data management and data governance in data fabric architectures. But even the people who are driving the data fabric narrative from an analysis perspective would tell you that we’re several years away from mainstream, from the data fabric perspective, and there’s probably only a handful of companies on the entire globe that have actually implemented a true data fabric, because data fabric requires some sort of semantic layer.

It requires some form of data quality layer, it requires some form of integration layer. It, it requires some form of active metadata where the metadata itself is informing the classification of data, the governance of data, the management of data, the integration and quality of data. So all these things are intrinsically intertwined and simply virtualizing your data, right? Enabling more modern integration patterns is not a data fabric, right?

All of those other pieces need to be put into play, including MDM, to enable a true data fabric architecture. Exciting stuff coming down the road. Happy to take off on conversations related to data fabric. But I think we’re a few years away from mainstream.

All right. So if anyone wants to debate that with Malcolm, go to LinkedIn and like, I’m there all day. I’m there all day. We have only five minutes.

So I to pivot to our last topic, but I really have to ask now, given all that conversation around data, fabric and data mesh, Patrick, is that something that you’re even talking about in Bank of Oklahoma or is that too far away to be practical? You know, I wouldn’t say that it’s too far away to be practical. But, you know, for my next three year strategy, I’m not trying to fit in an industry term. I’m trying to solve business problems.

So whether we end up looking a little like a mesh or a little like this or a little like that, I am constantly learning, trying to see where the trends are headed. But I’m not going to fit my strategy into anything that the industry says. This is what’s next, right? I’m going to take a list of challenges, list of corporate challenges, and I’m going to take I’m going to try to be about 2 or three years ahead in my strategy and not look much further beyond that.

I’m not saying that I’m going to take that I’m taking a shoot from the hip strategy. But if you try to go all in on any one thing, you’re probably not going to have the maturity to get there or the time or the budget. So I just go back to what I’ve been saying. I’ll take a list of I’ll start with my strategic priorities for my executive leadership team and now develop a roadmap that’s one to three years out based on where the technology is or almost at.

And then I go for that, right? I don’t want to implement something that’s obsolete when I implement it, but I’m not going to try to be the first one to do something. We’re not a fintech or we’re a bank, right? And we’re a large bank that’s going to become larger.

So we do that by making conservative sound decisions, but always paying attention to the future. So I will I’ll take a 1 2 three year view, I’ll take about a three year view and work backwards and implement solutions that fit. And probably it’s going to be very hybrid of any one particular industry concept. And maybe look back and say, well, this is actually what we’ve developed that looks a little like this or that.

Perfect oh, well, that sounds an awful lot like some practical advice for how to move forward. So let me ask Pedro the same question. Our final topic was how to operationalize and how to get going. Just given that we only have a very few minutes left, what would be one or two quick pieces of advice, Pedro, that you would give to anyone who’s looking to start this journey and build MDM and governance since their organization?

I think that one piece of advice I will give is understanding that it’s a team sport. The technologies alone cannot build this business alone cannot build this. It needs to have everybody putting effort and making it a cultural shift. These data governance initiatives and the it’s a change on the way that the company works is a cultural change.

So without having everybody involved, it is going to be a futile effort. What happens if you go with just technologies? becomes a database administration task. If you do it with just the business, then it just becomes about making money.

And then you completely forget all the other aspects of security and compliance and really like what keeps us safe on the technology space. Excellent Malcolm. Same question for you. Practical advice on how to get things going.

I know this is a topic that you’ve spent a lot of time talking about and on LinkedIn and all your various channels. Indeed, if there were two or three things, number one, find an executive stakeholder with very, very an acute pains who has motivated to solve for those pains. Sit down with that person and have a conversation about use cases that you need to solve and then quantify what the expected outcomes are through solving those use cases. Use very, very specific business terms.

Avoid the temptation to quantify or build a business case around data quality improvements. Instead, build a business case around more money, lower costs, or less risk. One of those three. Start simple.

Be focused on a specific outcome. Find a willing partner who will willing to work with you to stand something up, but keep your scope as absolutely limited as possible. If you’re focused on a limited set of outcomes, or even preferably just one outcome, that’s better. So start simple, be agile, don’t try to boil the ocean, manage for scope, and find the right business partner.

All right. I think everyone was nodding pretty, pretty violently there. So I think that spoke well for all of us. We’re at the top of the hour, so I’m going to actually stop it there, although I think we could potentially go on for a lot longer, but we will stop it here to respect everyone’s time.

I want to thank our panelists. Patrick from Bank of Oklahoma, Pedro from Odessa, and Malcolm, our own prophecy expert. I think everyone would agree there’s an awful lot of expertise in this conversation and a lot of words of wisdom and insight. So I Thank you all for your time.

And listen, Thank you, audience, for your time. And hopefully you got something out of this. If you’re interested in having a further conversation with prophecy, go to our homepage and there’s a Contact US button there where you can request demos or request conversation or whatsoever. Or as we said a couple of times, Malcolm is very active on LinkedIn and is starting a podcast, so feel free to jump in and all of that again.

Thanks, everyone, for your time. We appreciate it. And good day. Thanks, all. Thanks to everyone.