Ahead in the Cloud: Empowering Data Ownership with Sameli Mäenpää

Chad Watt: Welcome to the Ahead in the Cloud podcast where business leaders share what they've learned on their cloud journey. It's Data and Cloud Day here. I'm Chad Watt, Infosys Knowledge Institute researcher and writer, here today with Sameli Maenpaa, Chief Data Officer at OP Financial.

OP Financial is one of the largest banks in Finland. It has a 40% market share in banking and insurance markets. Sameli has been with OP for five years. Now Sameli, welcome and tell me how long has OP had a Chief Data Officer?

Sameli Maenpaa: It's actually pretty much the same five years I've been here because that position was the first time OP invented this kind of role and I think it's natural.

I mean, banking business is completely digital nowadays and most of the big local and global players, I mean, they are realizing that data is actually their number one asset right after the financial assets, of course, so I think it's a trend.

Chad Watt: Now before becoming the Chief Data Officer, you told me with a little bit with pride that you were a hardcore developer. Let's go back to those days and tell me how did you get started in coding?

Sameli Maenpaa: I always wanted to build something that people are happy to use, but I was not that good in building physical things. So I'm fortunate enough to have been born in an era where you can actually build intangible things. I mean, I'm good in problem solving, kind of understanding connections between things and talking to people.

I think software development is all about that. I mean, you need to talk to the people to understand what they need, what the customers need and what the employees need. That's how I sort of got started with that.

Chad Watt: You mentioned you were born at the right time to get into this kind of intangible kind of construction thing. What was your first computer?

Sameli Maenpaa: Amiga 500. That's an old one.

Chad Watt: Amiga 500. My goodness. That was color though. Yeah, I think you're a little younger than I am. Mine was a monochrome, I think.

Sameli Maenpaa: Yeah. Yeah. I mean I actually started coding ori with that around the time when I was maybe in second grade or something, but that was really basic things. I mean, nothing being fancy and I'm definitely not the sort of programming prodigy in that sense, but I started at least to get interested at that point of time already.

Chad Watt: Now talk to me a little bit about how you went from kind of coding and hardcore coding to managing systems and managing other coders.

Sameli Maenpaa: I've always wanted to make an impact, and most of us we do want to make an impact. There was a point in my career where I was thinking that, okay, as a coder, if I really want to make a bigger impact, I need to either found my own company and make impact as a coder through that company, or I actually have to be one of the best in the country or in the world. And I was thinking that maybe I'm not that good in coding and I was like thinking that maybe I should actually try to make a bigger impact by having many developers doing together with me the things that we believe are right for the people.

Chad Watt: Now, tell me a little bit more about your company for those outside of Finland. Tell us about OP Financial. What are you known for among the Finns and where does the name OP come from?

Sameli Maenpaa: Originally OP comes from the word [foreign language 00:03:27], and that's like if you directly translated it, it's like a co-op bank. And as we are a co-op, cooperative bank, I mean the origination is that there were some farmers who couldn't get from banks at that point of time 120 years ago, and they decided to join their forces and they created the co-op in that sense.

And since that, obviously things have evolved quite dramatically. And nowadays we have a 40% mortgage sharing both in banking and in insurance. So in Finland, we are very known. There is absolutely no one who wouldn't know OP, in Finland. We have a very strong brand. It is a kind of human-centric brand. I mean, the co-op history is very visible in our values and I think we are also considered to be one of the most valuable and trusted brands in Finland.

Chad Watt: Let's move on to some cloud calibration here. I know you're the Chief Data Officer, but data and cloud go hand in hand. What proportion of your data right now is in the cloud?

Sameli Maenpaa: When I did my MBA, I mean the professor always said that the right answer is, it depends, and it depends what you mean because we have the most important data assets in the cloud, but the same assets actually pretty much exist also in the on-premise environment.

And at the same time, you can say that you need to kind of separate the use to analytical use and operational use.

When you think about the operational use, very little operational use is in cloud. I mean by far, most of our processing and banking and insurance business happens sort of in the data center on premise.

But when it comes to the analytical use, almost all of that use is from the cloud, then we have very high coverage when it comes to that. And also I have to say that I mean not all data is equal. It's far from that. The most important assets, it's not the surprise if I say that financial account data, the payment data, credit card data, that kind of things, obviously, the core customer data, I mean that kind of assets, they post more value than some log data or some sort of minor part of the business data that we have, so we have pretty much prioritized which data we need to put into cloud first.

Chad Watt: Interesting. You want the most instructive data and perhaps more importantly, the applications and the analysis to be in the cloud. What can you do when you have that instructive data and the analytic tools in the cloud that perhaps you couldn't have done previously if those were either on premise or in different systems that you had to integrate?

Sameli Maenpaa: I think obviously cloud enables a lot of things that are not that easily available in on premise. I mean, many of the services like platforms as a service kind of things available in the cloud are accelerating the development significantly.

I think it is so important that you can sort of get into the development mode as fast as you can because in on premise, you spend a lot of time in creating the environments, the pipelines, and sometimes the scaling and so on, which is kind of native in cloud. I mean its kind of built for that. So it really helps in that, certainly, when you want to derive value from the data. Like I said, I mean data is not equal, so data is valuable actually only when you can connect that to the business and to the customers. So to me, data does not have intrinsic value and we derive the value more efficiently, I would say, in the cloud than in the on premise.

Chad Watt: That's very interesting. And this is something I've been working on for another project, and there was this analogy more than 15 years ago that data is the new oil and all you have to do is refine it and then you have value from that. And I've been toying with an analogy, not original to me, but that data is more like a nuclear fuel and that it has some value. You have to be careful with how you handle it and losing control of it is very risky. Give me a thumbs up or thumbs down on that. Is it data oil, is it nuclear fuel? Is it something else? Do you have something else you'd throw out there?

Sameli Maenpaa: Yeah, I mean I think it is oil. I think it is nuclear fuel, and I think it is gold, at least. So yeah, because I mean that's also something you can refine and especially the ore. I mean when you have the ore, I mean, it's not worth that much. I mean, you cannot really know how much gold you have if you have a pile of ore.

I mean you might have huge amount of sort sand and you have no idea if that's worth something or not. But when you refine it, then you sort of start to get a grasp of, okay, there is gold coming in. I mean one ton, two tons, three tons, and then you are very, very wealthy. So I think it doesn't matter. I mean whatever metaphor you use, the thing is that you have something raw and you have to refine it and then you have to use it.

I mean, refinement is kind of why would you do that? It's a cost if you cannot use it. So I think it is so important there to also understand how are you going to use the data? Are you going to reuse it? Are you going to sort of get some real value to the customer of the business out of it.

Chad Watt: OP, for your bank, for your insurance operation, is there anything in the data or app side that doesn't ultimately go to the cloud?

Sameli Maenpaa: Yeah, I mean obviously banks and insurance companies, I mean they are the pillars, one of the pillars of the society, and hence there are some regulatory and government requirements involved. I would say that because of those requirements, I don't think they will change their mind a hundred percent. It will always be sort some data and systems that have to be run kind of in an on-premise environment, which is very regulated. But I do believe innovation where most of the applications and data can be on cloud, both operational and analytical use.

Chad Watt: So we're talking about migrating to cloud in 2022. Can you give me some of your insight around what's different about migrating to cloud, adding things, applications and data to cloud now versus 2008, 2007 when Netflix is pioneering this cloud movement?

Sameli Maenpaa: So I think the culture, that's the first and foremost thing. I mean, I think in business, people are now having the awareness of the possibilities of the cloud. And it's almost like, even if you are a business leader who doesn't understand what cloud can offer, you are at least smart enough not to say that cloud is for nothing. I mean, you have heard enough speeches about that even though they don't understand the benefits, you are not that self-destructive that you would say that there is nothing in cloud and AI.

I mean you shouldn't invest in that at all. I mean that kind of business leader, they know that they would get fired pretty soon. I think the culture and awareness and understanding of the potential of cloud has made all the difference. But of course there are technical things as well.

So for example, the agility and scalability of cloud, that's insane. I mean, you cannot do that on premise in the same way. I always say that cloud is something that has brought democracy in the digital world. I think it is quite an important thing for the society that anyone can pretty much snap their fingers and get an environment and start development in half an hour.

You can get an instance and sort of run your code in half an hour, and it's quite different to the actual thing where you need to do an investment for the hardware. You have to sort of know what to do there and configure that and build the environments and so on. So the platform as a service kind of accelerators and also the cloud native tools, that really makes a difference also for sort of democracy I would say. But when it comes to banking, I think it applies also there that most certainly for the platforms as a service that we are using, they are giving efficiency to our operations and work.

Chad Watt: Since you raised the cloud native tools, can you, one or two really game changer, cloud native apps or tools that have helped you build the data driven organization you want to build?

Sameli Maenpaa: So I mean, I think one of the most important tools is Databricks, actually. That's been really a game changer for many because, before Databricks we had a situation where it was pretty much only the data scientist and kind of hardcore data analyst who could work with data.And now, when we have rolled out the Databricks and we have had a program called Citizen Data Science, which means that we are not going to sort of hire hundreds of more data scientists; we have like 30 plus. It's a decent amount, but what we need is to get almost as good data scientists out of the developers we have all across the organization.

And that's what it means to sort create the Citizen Data Science skill, as we call it, and I think we've been quite successful there. I mean, we have more than 300 Databricks users now. And obviously not all of them can build machine learning and kind of neural network based solutions, but many of them can actually build machine learning solutions that help the business. So I think it is so important to get tools like this, which are easy to use, which can be used by someone who is not maybe the hardcore engineer.

Chad Watt: I want to talk a little bit more about data driven by the cloud and there's an analogy I'll kind of lay out for you and then we can talk about how this applies in your strategy of building a data driven organization. There was a time when data was kind of output, data was the exhaust, it was kind of the byproduct of work and companies would use data to prove, well, that was a good idea, that was a poor idea, that was an idea we should consider revising.

Today, data is much more at the front end rather than the back end of the process. It's more instructive. Data used to be the exhaust, maybe it's the roadmap today. How do you use data upfront to build your organization? And you talk about data driven organization.

Sameli Maenpaa: Like I mentioned actually before, data does not really have an intrinsic value. I mean, it only has value if you can actually connect it to something real, the customer or the company's operations or something like that. So I think it is so important to get the understanding and of the culture, the support, the fact that data not only is an asset, but it is an asset that you have to treat the same way as the actual financial assets pretty much.

And I mean how I see it is that mean we have created even something we call data balance sheet where we have data assets and data liability is kind of the same way as in the financial balance sheet and I think it is so important that you can see the pile of ore, the pile of sand.

And if you don't know what's in it, I mean it might be even radioactive. I mean it might be dangerous, it might be polluted, but if you don't know what's in it, it's a liability. You need to know what's in it. And when you know what's in it, you can actually derive assets out of it. And that's how you do in a normal balance sheet that you have the raw capital and you derive assets out of that.

You buy machines, for example, or buildings. And that's on the asset side of the balance sheet. And I think it's the same here, the same analogies here that you need to put solutions onto the sort of left side of the balance sheet, make assets out of data. So I think that's one of the foundation of being data driven, that you can understand how that process works, how that value chain works.

Chad Watt: Let's talk about data, AI and cloud. Are you putting AI to work with your cloud and your data?

Sameli Maenpaa: Yeah, we have invested into AI for the past few years.

Chad Watt: How's that faring?

Sameli Maenpaa: I think we have actually achieved a lot. We started from almost zero like five years ago, and now we have more than 20 models in production and we actually have some huge cases where we can derive some real value. I mean, if I think about the value we have derived from AI and, you would say advanced robotics, it's something like a 100,000,000 roughly in benefits.

Chad Watt: I want to hear about some use cases from you. What's the most valuable use case and what's the smartest, most fun puzzle you've solved?

Sameli Maenpaa: I think the most valuable use case would have to be the one that we implemented. I mean, I think it was when the COVID started in 2020. That was sort of a time when a lot of people wanted to make changes into their payment terms and make sort of new schedules for payments and that kind of things. And normally, per day we got a 100, 200, 300 requests for that. We were happy to do some manual handling there and it was not the problem. But during these times and early days of COVID, we were in a situation, that amount of request was tenfold. So we got like 3,000, 4,000, 4,500 requests for changing payment terms and schedule. So that was a tough situation because if you are one of those people, I mean, maybe you lost your job, maybe you don't get salary anymore from your employer.

I mean, you have a real situation where it doesn't help you if the bank would say that. Yeah, but there are others also having the same situation, so you will be processed in due time. I mean, that doesn't really help. I mean, you are not satisfied. And we actually managed to do in a few weeks because we had some assets that we had done before. In a few weeks, we managed to do a solution with the help of AI and robotics where we managed to get, sort of, tenfold more efficiency out of the process.

That was really something that saved the backbone of the bank and I think especially it saved a lot of customers and I think that's even more important than the backbone of the bank. So that's probably the biggest sort of success story. And I mean, every time I see the CEO, he's always saying that. Can we do more of those?

Chad Watt: Well, thank you Sameli very much for your time and insights today.

Sameli Maenpaa: Thank you. Nice chat.

Chad Watt: This podcast is part of our collaboration with MIT Tech Review in partnership with Infosys Cobalt. Visit our content hub on technologyreview.com to learn more about how businesses across the globe are moving from cloud chaos to cloud clarity. Be sure to follow Ahead in the Cloud wherever you get your podcast. You can find more details in our show notes and transcripts at infosys.com/IKI, that's in our podcast section. Thanks to our producers, Catherine Burdette, Christine Calhoun, and Yulia De Bari. Dode Bigley is our audio technician and I'm Chad Watt with the Infosys Knowledge Institute. Until next time, keep learning and keep sharing.