Organizations are experiencing significant data sprawl. It is getting complex to get accurate insights timely from the data one has in hand, or it is lying there unused. What are the gaps in our data strategy, architecture, and execution processes? How are we modernizing data management to facilitate seamless access, sharing, and processing in a distributed environment across applications, systems, and platforms to support business outcomes?

Contributor

• • Luis Fernando Lopez, Chief Data & Analytics Officer LATAM, Chubb
  • Navi Grewal, Global Chief Digital and Information Officer, MasterBrand Cabinets, Inc
  • Ravi Krishnan, Chief Data & Analytics Officer, State of North Dakota

Download Podcast

Apple Podcast, Google Podcast, Spotify, iHeartRadio, SoundCloud, TuneIn, and Stitcher. Find other syndication channels here or search CIO Talk Network podcast on any other app.

Explore More

• • Data Management: If the shoe fits, then let’s run
  • Data Management: The Loch Ness Monster or The Boogeyman?
  • Creating a Winning Cyberthreat/Ransomware Management Strategy
  • Digital-Enabled Wealth Management
  • Creating a Winning Data and Analytics Strategy
  • Addressing The Big Data Veracity Challenge
  • Building Digital Data Strategy, Architecture, and Infrastructure

Transcript

Sanjog Aul [00:00:00]:
Hello and welcome to CTN. To learn more about the show, please visit ciotalknetwork.com and today’s topic is Modernizing Data Management. What are we talking here? So we have an issue with data sprawl among most organizations. A whole lot of data coming from all different directions, and we have to manage it. So when we talk about that, it is getting very complex and at the same time, you’ve been mandated as leaders to get some very accurate insights, timely, which could be handed over to the right constituents so that they can make use of it and create business value. That’s a problem and then there is a lot of data which nobody knows exists, which could, if somebody had known, have utilized for business purposes.

Sanjog Aul [00:00:49]:
So we have all these problems that we are grappling with. So idea that we wanted to discuss on this forum is primarily first to inventory the gaps that exist in our data strategy and architecture and even the execution process of this whole data management which could be causing these issues. Then let’s talk about the potential ways we can modernize this data management discipline so that the data that is available, it is available seamlessly and accessible seamlessly to all people who require it. It can be easily shared and it can be processed in a distributed environment across applications, across systems and across platforms so that business gets what truly wants by exploiting the very value of data, which is lately the currency for digital. So to discuss this, we have Luis Fernando Lopez, who’s the Chief Data and Analytics Officer, LATAM for Chubb. Hey, Luis, how are you?

Luis Fernando Lopez [00:01:52]:
I’m good. Thank you for inviting me.

Sanjog Aul [00:01:54]:
Great to have you sir and we have Navi Grewal, Global Chief Digital and Information Officer with MasterBrand Cabinets, Inc. Hey, Navi, how’s life?

Navi Grewal [00:02:02]:
All good. Hey, Sanjay.

Sanjog Aul [00:02:04]:
Great and we have Ravi Krishnan, Chief Data and Analytics Officer with the State of North Dakota. Hey, Ravi. How are things?

Ravi Krishnan [00:02:12]:
Things are great. Good morning, everybody.

Sanjog Aul [00:02:15]:
Awesome. So, yeah, let’s start with you, Luis. I’ve literally been talking about data with leaders like yourself for the last 19 years and I’ve always heard and yes, of course, in the last decade, this noise about data increased, that data sprawl is increasing. But then people also invested a whole lot of money, time, energy into this problem but it seems like the inefficiencies, the cluginess, the complexity doesn’t seem to go away. What are we missing?

Luis Fernando Lopez [00:02:49]:
I think we are. We’re living with a lot of purpose. Right. We’re trying to get data as soon as we can. Right. We, the organizations have been with have switched the view from trying to do the bare minimum of getting sales dashboards and trying to get data to make decisions in a very use case dedicated way and very small part and then for all of a sudden we started moving to big data and trying to get everything even if we don’t know what do we want to use and that’s where the problem actually lies because we’re, we switched this view we try to change from I’m going to take this, we’ll have might help me make better decisions too.

Luis Fernando Lopez [00:03:34]:
I’m gonna have it all and see if some of this makes sense and see if some of this actually gets in the way of me getting a new insight, something that was never discovered and the possibility of that in my opinion is very narrow. Like the chances that you are getting an insight that people with 20, 25 years in the business does not know, does not thing to grasp.

Luis Fernando Lopez [00:03:59]:
It’s very little.

Luis Fernando Lopez [00:04:01]:
It’s definitely possible, but I think we have to start going a little bit back to actually getting permission that we’re going to use. The problem with this is, our cloud servers are getting bigger and bigger every day and our dashboards are getting more info and they’re becoming unmanageable. We have to be more lean about the information we get. We have to get more lean. We have to work with the business and say, hey, what are you actually going to use? If I get this much information, are you going to really use it to make decisions or are you just going to ignore it and let it stand there and make decisions with what you already do and that way we can be honest with our companies, with our expenses, with our culture and really make data driven decisions but with data that we are actually going to use, I think that’s part of the bigger problem. We’re trying to make much, much more

Luis Fernando Lopez [00:04:56]:
and we’re lost in that translation of getting much more.

Sanjog Aul [00:05:01]:
So Luis, based on what you said, it seems to be like a data strategy development problem because we don’t even have a blueprint on how we want to handle data. So to that end, Navi, coming to you. If data strategy is not made relevant or it’s not staying relevant, and basically when it’s not staying relevant means whatever is happening at the business level, perhaps data strategy is not able to catch up or maybe it was never created as aligned with the business environment. So what’s preventing it? Are we like to Luis comment there that are we always thinking too narrow or we are rolling the dice to say okay, let’s get a bunch of data from wherever possible and then see if it would help and it’s almost like rolling a dice. Is that what’s happening?

Navi Grewal [00:05:51]:
I feel, Sanjog, that our approach sometimes is faulty, right? We focus too much on data strategy. For me, we should be changing the paradigm to information strategy, right? Start with the question, what is the business trying to solve for and over the years, it’s evolved. Pre Covid, there was a lot of focus on consumer and I want to understand the segments. During Covid it cordially switched the supply chain. What is the pricing, can I catch up with pricing. Now I think the focus in the next couple of years is going to shift to inventory. Next is going to be now. Everyone wanted everything quickly and now everyone’s going to be based on how the economy is going to go. So to me, for us to be relevant with the data strategy, what we have to do is be in long step with the business to say, what are the problems that you’re going to try to solve in the future and then go back and think about, what is the data do I need? Where does the data come from? I think we sometimes get too hung up on the data and with ERP and all that comes from and we stop the focus on the information.

Navi Grewal [00:06:58]:
The other thing is the trust in the data. I think the biggest thing in the business is that when you give them the data, they don’t trust the data. So I think the approach to say, okay, let’s start with the minimal viable and tell us where you see the data issue because what happens a lot of times is they were like, oh, I’m not sure if the data is 100% accurate Navi, what am I going to do is download the data into Access or Excel or SQL databases and now try to change the data to answer the questions in the way that I want it answered. So I think what we have to do as a leadership team is stop that and say, god, we’ve got to fix the issue and let’s go back to the source and fix it, rather than us having multiple offline sources, which then we lose control and then that leads to this whole data strategy being a failure. So those would be the two things that I would focus on.

Navi Grewal [00:07:55]:
Sanjog.

Sanjog Aul [00:07:57]:
So, Ravi, based on what Navi mentioned, that, as a business you have to look at business first and what does the business want and that, of course, is Leadership 101 and someone is missing there because of which we are having these Issues and I’d like to bring up this point that Chief Data officer role, of course was introduced almost, I would say five to seven to ten years ago now for the simple reason that they needed some specific dedicated effort to that end where data is seen as an asset and it’s not just floating around and we’re trying to see if we can make some sense of it. They are made owners and other stewards of this very valuable asset. Even though we did that, what do you think we missed out? Even though you’ve been given this dedicated role or any data officer is supposed to work on, what is preventing you from making this all come together and we sing Kumbaya and life is beautiful?

Ravi Krishnan [00:08:55]:
That’s a great question, Sanjog. The Chief Data officer in any organization is almost expected to be like a magician. Next comes in with a magic wand that’s part of the toolkit and they just wave it around, get things done and solved. The reality is there is not one type of Chief Data Officer. So there’s a Chief Data Officer, there’s a Chief Data and Analytics Officer. Some companies actually separate those roles. They put Data Officer in the IT side, Analytics Officer on the business side. Some have a different combination

Ravi Krishnan [00:09:28]:
but the bulk of the thesis is this, right. Data as an asset is not owned by IT. Data as an asset should be owned by the business and that is the primary role of the Chief Data Officer, to bring that awareness that ownership and governance has to belong with the business. We can enable you within the IT side. We can provide the guardrails, we can provide the tools, the technologies, the know-how, and so on and so forth but the real work of data lives in the business. When I say data ownership, it means all kinds of things, right? How data is sourced, where the data is obtained, from which vendor, how do we get the good quality data from these vendors, internal or external sources, and also to make sure that the data is trustworthy. Can the business use the data to make those decisions that they’re looking to do? So I think that is the big piece which a lot of organizations tend to miss and I have seen in my experience.

Ravi Krishnan [00:10:26]:
More recently, I’m in the public sector, but I have spent a good chunk of my career in the private sector running large data strategy, data governance engagements. So in my mind, unless data strategy and governance are clubbed together, John was joined at the hip and said we got to do data strategy, very important, but equally important is data governance, to put those structures in place. We call a business data steward, there’s got to be a corresponding technical data steward as an example. So they have counterparts in that journey and that way we can ensure that whatever we are trying to do in terms of business outcomes, answering those business questions, getting those business solutions out there are all in lockstep. So that in my mind is the biggest hurdle to cross.

Ravi Krishnan [00:11:12]:
It’s a hard one because this involves all kinds of things. It requires a culture change, the organization culture, people are supporting. It requires data literacy. Everybody needs to be aware of what data is, how it can be used, what to do with it, and so on and so forth and then the execution part of it. You need people that have been there, done that, being able to show the ropes to others with less experience. Many data programs fall to because we aren’t putting these control structures in place at the outset. We do it as an afterthought.

Ravi Krishnan [00:11:44]:
So in my mind that makes the big difference.

Navi Grewal [00:11:48]:
I just wanted to add Sanjog, so I totally agree with you Ravi and I think I saw the evolution to your point where we started with enterprise data warehouses, our data leakes. So now we’re moving more towards a domain based architecture, where each domain is owned by harder product and they are tasked with organizing the datasets a lot and to me, rather than one organization or a central team owning everything and not understanding the data really well and handing it over to the product owners who can actually go in, educate people about what does the data mean, what kind of insights can we drive, has worked really well where we have implemented that kind of org structure.

Ravi Krishnan [00:12:28]:
Oh absolutely.

Luis Fernando Lopez [00:12:31]:
Sorry, I just want to add. I’m pretty sure I’m not alone here. I think we’re in a very complicated spot where things, a lot of things were not done at all in the past. I’ve worked in companies that spent 15 or 12 years with no data work at all and right now we’re expecting students to be able to deliver big machine learning and AI algorithms that run near real time and they’re going to be game changers for the business. We don’t even have foundations, we don’t even have word to run those algorithms.

Luis Fernando Lopez [00:13:14]:
You’re like you want to talk about and we have had in other companies, business leaders that go hey, we’re going to do AI and we’re going to do this much and in Ford’s magazine, we’re not delivering that, we’re not doing that. It started to be for some part of the business thing to brag about to say we’re doing this, but we’re actually very behind trying to do that. I don’t know if you get that feeling too.

Sanjog Aul [00:13:47]:
So, Luis, based on the comments you made, Ravi made and Navi made, I’m going to go to say your organization, Chubb, it thrives on data, right? So you essentially use that as a currency to do literally everything. It’s a services organization. It’s very data heavy and you were brought in, I’m not sure how long ago, but whosoever worked on data, I’m sure at the business level, I assume that they would be enlightened that if we did not manage data, well, God bless us all, right? And you might have tried to do things to make it better. So is the intent missing from a business side to take ownership and say, okay, we’ll put all our might into fixing data, or the ability is missing, or it is pure inertia which is holding the organization? I’m not sure what your current state, so it’ll be good for us to learn or even overall, as other organizations which are very dependent on data, is it the intent that is missing, is it the ability that is missing? Or if there’s a pure inertia, then why bother?

Luis Fernando Lopez [00:14:53]:
I would go with inertia. As I was trying to express before, right now in chop, there’s a kind of a sweet spot. We don’t lack data delivery. We don’t lack data, we don’t lack will, which is very important. Like the business wants to do this, it strives to do. It’s a problem in some companies that they’re not so open but in Java, we are willing.

Luis Fernando Lopez [00:15:18]:
We have data, we’re willing to use it, we have the right resources in place. The thing is, we realized we needed this, kind of late, when everything started going so fast, then data stopped at a point and then we need data to leap from where it is to where we want it to be, to make more automated decisions, to have underwriters focus on getting more work or more customers instead of trying to sign the past customers. We need data to leap from one place to the other and there is not enough hours on the day to do that. So that’s the problem. We haven’t found the rhythm. The other three pillars for me that are, we have data or we know we have it, business wants to do it and we have the right resources are fundamental to this operation to work like, we just have to get the rhythm in place.

Luis Fernando Lopez [00:16:16]:
We have to go to a better pace, but not literally on me. This is the first time overall that I have the tools in the right place and the people in the right place and it feels way different to work in this way than the others.

Sanjog Aul [00:16:31]:
So you’re saying it is more like a project now versus a problem. You just get it done and next year I come to you say, beautiful, everything is in place.

Luis Fernando Lopez [00:16:38]:
Well, next year, because it’s just around the corner. See, in 12 months we can talk.

Sanjog Aul [00:16:47]:
Okay, beautiful. Now, to that end, now we’re coming to you. Since we heard inertia as one of the things and we have had some other issues that were raised, would you say that it’s not truly a problem, it is something that we are all learning? We are evolving, we are improvising, and that is how we will get a reasonably good handle. We will never reach a perfect stage, but good enough would be a good thing to aspire for when it comes to data management.

Navi Grewal [00:17:19]:
Yeah. So I would say there isn’t a world that would ever be perfect anytime. With the amount of data coming in at us, the variety of data coming in at us, and the business needs for the data evolving will always be good enough and then we’ll jump to another thing. Going back to Luis’s point, I think one thing that I’ve seen always as a roadblock from a data strategy and analytics perspective, is the speed. Always, being in IT, like hiring the businesses won’t come to us because IT takes forever because we try to look for the perfection. I want that data coming in and it was perfect.

Navi Grewal [00:17:58]:
And when they come to us and I mean, I want to know where is my order and I’ll tell them, “Guys I need this erp, that erp”, I have so many backend systems, it’s going to take me four months to get back to you the link. Four months? By the time my business problem is solved and you are coming in four months? So I think that’s where we have to focus on and what is our approach? How do we add the business value? Earlier we took a lot of the ‘the waterfall approach’ to say everything has to be perfect.

Navi Grewal [00:18:28]:
Then I’ll give you stuff at the end, I think moving more towards the Agile framework to say, let’s solve the business problem iteratively. Let’s give you some data and then add on to that so that there are every two weeks sprint and people see their value. I think speed has always been the issue for everyone in the BU’s and the functions to come to us and that’s something that we have to solve. I don’t think it’s as much about the talent. It’s just our approach to adding business value and solving their problems has been a big challenge.

Sanjog Aul [00:19:03]:
So any specific areas that you recommend people start paying a close attention to, which is whether in terms of data technology, business architecture, where if things were improved, you will get a quantum leap ahead in this whole journey.

Navi Grewal [00:19:20]:
I think, to be honest Sanjog, as I said I think it’s the approach, how do we implement, how do we approach? I think more of agile parts where we come up with- who is the product owner, what is the business problem we’re solving, trying to focus on the developers, and so that’s a part that’s working on a business problem, then we can spin out multiple parts at the same time. So to me, if we can focus on the execution and then the other approach that we’ve taken is the platform and the product approach. There are engineers, there are people in the backend that are just getting that business demand and they’re churning. So you stand up the data factory in a way that the demand comes in, you’re able to get the output in a much quicker fashion. So to me, setting up those pods to work in parallel when the product owner knows what business problem is, what the data means, and then having that engineering and data engineer and the developer layer that takes the demand, I think that’s the way for us to solve that speed issue and the problem. That to me will be the one that adds the most value but I would love to hear Ravi and Luis’s approach to that.

Sanjog Aul [00:20:28]:
Let’s take a quick break, listeners. We’ll be right back and when we come back, Ravi, I would love for you to respond to Navi’s question and also let’s talk about the execution challenges which we may be facing, which is causing us to be in the state that we are in. It’s not a bad state, but I think we can do a lot better but please stay tuned listeners. We’ll be right back.

Sanjog Aul [00:21:49]:
Welcome back. Ravi, Navi asked you for your input and also once you shared your thoughts on her points that she made and she very well allowed a segue into the execution and my next question to you will be about the execution challenges and that we are facing which is preventing us from reaching at least a good enough state.

Ravi Krishnan [00:22:13]:
Thank you for that question Sanjog. Navi, good points there that you made. I am just fresh off a Gardbeer conference this week. Last week we had well over 6,000 people show up for the first time in three years. So it is refreshing to see a sea of people and they’re all data practitioners, data gurus, CDOs and so on. To me it feels like through this pandemic something has changed and I want to start with this right? In the good old days we had a simple data warehouse. There’s a Kimball warehouse, a nimble warehouse, those data mods.

Ravi Krishnan [00:22:48]:
Through this pandemic you realize not four people have done data lakes. A few had done it before. So we have data lakes on the cloud, right? Then a few people tried data meshes and now the in team item is data fabric. Everybody in the family is doing data fabric, right? So we have a multi cloud architecture with data fabric and a lot of the experts are also pushing this view. So this is the challenge that the CDO faces, right? Working with his or her counterparts in technology, in security, in business to think about how do we future proof our architecture, whether it’s a technical architecture or business architecture. As a result, you know what happens is many organizations end up going into POC mode, right? So they’re continuously doing POC. So we have part 11 doing POC for XYZ.

Ravi Krishnan [00:23:41]:
Part 12 is trying something else. So we are in a race among ourselves to figure out which is the best architecture, the best way to move forward. Some organizations that are more conservative take the less risky approaches. Say we have a working architecture, we have a data lake, we have a data warehouse. Let’s not break it, let’s just keep going but if you want true transformation, you have to mix some of these more cutting edge technologies so that in five years you’re ready with the more current priorities. Which leads me to my primary point here, which is around competing priorities. The biggest challenge that we face today is the lack of clear prioritization.

Ravi Krishnan [00:24:20]:
The business will tell you we have 14 items that are number one priority. So how do you prioritize and some agencies, some organizations have three, four, five lines of business. In state government, I have 75 agencies I work with, so we have 75 number ones, right? So who do you pick, where do you start, where do you execute and who gets the bulk of your attention? I think that’s a big question to be answered by IT leadership, technically where the data organization lives but it could be on the business side also. It really has to be driven in such a way that we’re able to deliver on the top two or three items. Anything more than two or three top priorities and we’re setting ourselves for failure upfront. Now, in the execution side of things, I agree with Navi’s points on our age of transformation.

Ravi Krishnan [00:25:10]:
We’ve all done that. We have all brought in angel coaches by the bus load, right? They’ll drop ship into your organization, they’ll go sit in their pods, they’ll reach the glories of agile and scrum and Kanban until the cows come home and after a while, for 6-12 months, everything is fantastic. Everybody’s still drinking their Agile Kool Aid. When you hit 12 months, that’s when something changes. Why are you doing these? Why do we need a retrospective every two weeks? Why do you need to sit down? Or continuous improvement every week or every two weeks.

Ravi Krishnan [00:25:44]:
So teams start having that age of fatigue. This is where the leader steps in and says, you know what, this is great. We gave you a good framework to start with, but we want you to improvise. We want you to customize your approach in a way that meets not just the needs of your customers, but also keeps your employees happy, your morale high. So that ability to twist, to pivot, to adjust, I think it’s going to be crucial for execution because the motivation levels of employees through this pandemic and with a great resignation, we’ve all been grappling with that. Resignations have been flying in left and right. There’s really nothing you can do. Just sit back and watch in many cases. So we have to be making sure that we build an environment that is more employee friendly, that is supportive of the work that they’re doing.

Ravi Krishnan [00:26:32]:
And in a way enabling that execution that we need for the business. So those are areas that I would focus on, which is making sure that we are building that clean environment where everybody is clear on the priorities, nothing is behind a dark wall, everything is out in the open, it’s transparent, we know what our priorities are for the next year and then giving ourselves the freedom to experiment, giving the teams a space to go and execute, I think those are the keys to success.

Luis Fernando Lopez [00:26:59]:
I just wanted to get into two points of Ravi’s participation. The two are mixed together because I think this data leak strategy, there are new things coming and there are new technologies, new ways of putting data together and then the leap from data warehouse was big or not so big to what’s coming and we haven’t even talked about blockchain and stuff but the thing is, when we finish delivering a full 100% state of the art daily, things are going to be past, move daily.

Luis Fernando Lopez [00:27:41]:
That’s the thing that worries me because we’re moving at a pace globally. Data management strategies are moving at a pace that most companies cannot move. I think companies have to stick to say, we’re going to do this and even if in five years everyone changed their data management strategies to blockchain, we’re going to stay here because we’re going to do a little bit of this and a little bit of that and a little bit of other things because we’re not going to go mad and this data problem we’ve talked about before is going to be crazy. If we have different technologies right now. We have three different clouds.

Luis Fernando Lopez [00:28:21]:
Then another one can be coming, another one can be challenging the way we haven’t and we’re going to be dividing information more. The same point to Agile methodology, because I have found out that Agile doesn’t really work for my teams. Like, it’s not a work that you can be specific saying, hey, how much work are we going to take in just in this two weeks? No, it’s not going to take two weeks, it’s going to take more. Okay, then we’re going to take three sprints to ingest this. Then the whole purpose of it changes.

Luis Fernando Lopez [00:28:56]:
There is no methodology that sticks to data as hard as we want it and we have to be realistic. One, we have to stick to the strategy for the long term and two, we have to find a methodology that works for everyone. I literally took a step back and said, we’re going to go with example, we’re going to go with a backlog and say how much are you going to take to do this and what are your roadmaps? I’m sorry because we take a lot of time doing retrospectives and planning screens and doing this and the first deliverable is not even reached. Like we take too long to deliver stuff. Two weeks sprints doesn’t work.

Navi Grewal [00:29:41]:
I mean just going back to the point of adrenaline and the sprint. So actually it really ravel. There’s like when you change the culture, you put in adrenaline and everything that stick forever, everything is changing, evolving on a continuous rate. I would just give you a simple example. The biggest issue that our company or from the customer we heard is, Navi I don’t know where my order is and then we ship the kitchen and cabinets but some people keep redoing the kitchen or beating the house, right. So when somebody places an order all the way when it gets shipped to a customer, there are probably 50 data points that it stops at.

Navi Grewal [00:30:22]:
So my thought was, let me focus on agile and say I can give you three stops. So let’s start with the customer to say you place the order, we received it and now it’s on the truck and gone and then we add in more data points for them to understand. So there it took us six months to get to where my order is, but then we took a lot of agile and then after that, anything beyond that, it worsens. So I totally understand not everything works in an agile manner. Not every culture is ready for that. So you have to make sure it’s going to stick and it’s going to be ready for your culture. One question that I had for Ravi, I know you talked about this innovation and new things coming in.

Navi Grewal [00:31:03]:
How is your team organized? Do you have a run and maintain team which is like dealing with the business requests coming in because the business still has to run and there are new requests coming in and there is all these innovative ideas. I think every vision is something new that comes first of all. You have to see which is noise and what is real. Do you have a team that’s focused on and is ready then you implement it in mainstream? Because that’s something that I am taking through two. Should I have a separate team there, run and maintain and how do I balance that?

Ravi Krishnan [00:31:40]:
Good feedback there, Navi, and good question as well. We have what we call a work intake team right across the organization and this is composed of members from every vertical. We call them the Chief Data Officer, Vertical Chief Technology Officer, Vertical Chief Security Officer, Chief Customer Success. So most organizations have some form of work intake, some kind of backlog, grooming, prioritization. My experience tells me it’s always a work in progress, it’s never perfected. Every organization I’ve worked at struggling with this whole concept of work intake.

Ravi Krishnan [00:32:15]:
How do you prioritize which one is more important, who gets to work on it but we’ve gotten it down to a point where we’re able to say when business assigns any, pushed all our business users to say we want to use one of the tools. I will name the tool Sanju. Just so we are tooling us have a tool that business can actually start entering ideas, right? This is what I like explore. This is my budget, this is what I want to do. So that comes into a central pool where we sit down as leaders and evaluate the requests and we start funneling it and triaging it to different divisions. Somebody needs to go research it. Enterprise architecture, please go weigh in on this. Is this the right approach? What tools do we need? How do we build this and we have that model where a lot of the waiting out occurs before we even write a single line of code and before we send it to any team.

Ravi Krishnan [00:33:04]:
But once we are all aligned, this is when we send it to the agile teams for execution. So within the CDO area we have 22 agile teams. We call them pods or Agile teams and SCRUM teams or whatever you like. So each of these 22 teams, they do the entire gamut of work. So My responsibilities include ERP systems, so we got PeopleSoft systems. We have a lot of developers also, which is interesting for our chief officer. So we have .NET developers and Java developers. So I’m able to catch where the data is being generated by these applications.

Ravi Krishnan [00:33:38]:
We also have a lot of the traditional DNA data analytics functions, data management functions, business intelligence, data warehousing, cloud and data scientists. So we are one of the few state governments in the country that actually has data scientists on our payroll doing active data science and machine learning projects. So we’re very proud of that and in the middle of all this we also have what we call geospatial information systems (GIS), a lot of the map data, weather data, traffic data, snowstorms and so on. We can get to have a nice product overall or spread of all these Data elements being created. So it gives us a nice window to organize it well and our ultimate goal is to be able to connect these teams as feeder teams.

Ravi Krishnan [00:34:22]:
So let’s just say a data warehouse team builds a warehouse. The business intelligence team builds dashboards on top of it and the data science team is then able to feed off the work of these teams. So I want to create almost like a little bit of a supply chain internally. It gets complicated very quickly when you apply things like a scaled agile framework. Many of you have experienced this scrum at scale agile framework where you have the concept of agile release trains. So within an entire organization, you think of this as a train with an engine and compartments, right? Each of these compartments represents one agile team, and they all sit and plan together every three months.

Ravi Krishnan [00:35:01]:
But where it gets complicated is when you have multiple trains running. The Chief Technology Officer may say, you know what, I need my own training, so let me build my own age of teams. So first we’re talking hundreds of agile teams across the organization. Literally in the hundreds and how do you keep them in sync? This is where the role of a traffic cop comes in. Traffic cop who trains for signaling mechanism. So this train comes and delivers business value till this point. This is where the handoff occurs to this next train. You pick it up, you execute.

Ravi Krishnan [00:35:32]:
So this is where a lot of that coordination among the leadership teams is very important to have. I have seen this work really well in the private sector. In government, it’s a bigger challenge because more than anything else, we are budget constrained. So we are all funded by taxpayer dollars. So we’re very sensitive to how we spend but in the private sector, I have seen that we were able to execute things like this at a much grander scale because of our “unlimited budgets”, but those would be some ideas for execution. I’m curious to see how your areas are organized, Navi and Luis.

Sanjog Aul [00:36:10]:
All right, so let’s take a quick break, listeners. We’ll be right back after these messages and thanks so much, Ravi, for your input and when we come back, Luis, starting with you, love to get your feedback on what Ravi mentioned as a roadmap, or rather a blueprint of how you can align teams and where do you stand with respect to that if it was to be used as a benchmark? So please stay tuned. We’ll be right back.

Sanjog Aul [00:37:31]:
Welcome back. Luis. Coming to you based on what Ravi has laid out, that is a pretty sophisticated approach to getting different trains running and signaling and orchestrating. What do you think are the issues that you feel, Luis, that will hold you back from building this ideal blueprint or executing on?

Luis Fernando Lopez [00:37:52]:
I think you get too lost. I want to say I’m execution first. So you need to get deliverables. Deliverables are not met from little. It’s going to be going to be really challenging. The first thing is you have to understand what are the tactical solutions that need to be in order to build trust and to earn some time and deliver those. Those are the main things. You have to focus on two or three projects and say, these are not gonna iron it up. All get your focus full in and build your strategy for five years or 10 years or how long you want.

Luis Fernando Lopez [00:38:34]:
So if you deliver those, in two or three, that are more complicated and deliver those and once you have delivered those 5, 6, 10 projects, then you’re going to have earned trust, you’re going to have earned benefits, you’re going to have earned some real sponsors to back your strategy saying, this is actually working and it can be in a full length of data, it can be a dashboard, it can be a database that everyone’s doing manually once a week. It can be an algorithm that actually does AI or some piece of the work. They don’t have to be real complicated things, they have to be things and this is the most important part for me is I always talk live streams. I always try to teach them this. I work for insurance. I am not hired to do state of the art data. I am hired to deliver data for business makers to make better decisions.

Luis Fernando Lopez [00:39:37]:
I am not getting paid to challenge Amazon or challenge Netflix in their architecture or in the pace they do. I am getting paid for them to do better than we do right now and to make the insurance industry better if I can reach. That’s a good point for me. If I try to say, we’re going to be the Netflix insurance, the gap between us is too big. I wouldn’t know where to start. So little by little start building your data, start delivering projects. Then you can do things that are in the long term.

Luis Fernando Lopez [00:40:14]:
But otherwise you’re just going to stay in between. You never delivered anything and you’re far too deep in the project to change and then the harder conversations start and I think we’ve all made that mistake saying hey, we’re not going to do any tactical, we’re going to go endgame and the end game didn’t work out. In the end you have to build tactical, you have to build small solutions that help people and that make them understand that the end game is great but execution is first.

Sanjog Aul [00:40:50]:
Navi. So you have any thoughts on what you think has happened at your end and what’s worked, what’s not worked, as you try to build this beautiful blueprint and bring it to execution?

Navi Grewal [00:41:00]:
Yeah, I think one of the things that have worked really well, at least for our company, the adult approach because of the speed at which we can deliver. The second thing that we’ve done is we’ve started to hire product owner. So earlier that has been one of the my missions, that data is the data strategy. There is a governance element that should be owned centrally by the Digital or the Information Officer but then there are domains and ownership and how the data gets used and all that needs to reside in all the business units and within the function.

Navi Grewal [00:41:36]:
So we need a data evangelist in each of those groups so that they can take work very closely with us and deliver the business value that we’re looking for. One of the a couple of areas where we have hired really strong product owners doing like in our sales and marketing. Those data insights and those dashboards, they are top notch. That person is evolving and continuously improving them and giving more and more information that we need and works very closely. It’s almost an extended part of your team. Now with COVID supply chain was another issue. With a lot of companies, where is the product, are we getting it, where is the inflation, where is my order, where is all the raw materials?

Navi Grewal [00:42:19]:
So now we’ve hired a supply chain data leader and that is working very well with our team then now we’re making way faster and more progress in that area because we have somebody who working with the teams to push that agenda forward. So to me one of that org structures where there is central teams and then there are business owner and the functional teams are really, really working for us.

Sanjog Aul [00:42:46]:
So Ravi, let me come to you about like you mentioned that you got great execution, at least the blueprint is there and you’re working towards essentially executing the best which you can but while all along you do that, people are seriously concerned about security and the governance of something that beautiful that you put together because if it is not, somebody will get fired or lose their shirt if there’s a breach and also if it is not governed properly, what you created as a good looking castle could come crumbling down very quickly if governance were not put in place. So where do you think alongside that beautiful blueprint that you mentioned, we could safeguard from a security and governance standpoint or what are you doing to that end?

Ravi Krishnan [00:43:34]:
Very important point and question here, I’ll say. One of the most important relationships that a Chief Data Officer needs to have is with the head of security, the Chief Information Security Officer, CISO is what we call in many companies the head of compliance and the head of governance. Depending on how we put it, it becomes very important, especially in companies that are more financial services oriented. So you have SOCs, you have basil regulations in the pharmaceutical industry, you got other regulations to deal with healthcare as well. So every industry is slightly different. So we gotta be aware of the regulations and even though the it is not a very new and exciting area for a chief lead officer who rather talk Machine Learning and AI all day, it’s very important for the leader of the organization to be aware of what is GDPR, what is CCPA, why do I care? So having that education upfront and being able to tell that to your teams I think is very important.

Ravi Krishnan [00:44:33]:
The more the Chief Data Officer talks about these things, the more that people become part of the D and E, the culture. So that’s very important. The second aspect I’ll say is when we start these projects, when we have these massive strategy engagements or invocation engagements, it’s important to involve some of these folks up front in the beginning, right? When you’re doing requirements, you’re trying to figure out what this is going to look like. So having secure, giving security a seat at the table I think is really important. One, they feel valued and two, when you actually are ready to go impress the business, they won’t shut you down. I have seen numerous examples in my career where CDOs have done an amazing job of corralling the groups, getting everything done, only to be shut down at the very last minute and they couldn’t make the go live dates and promises to the business because security said you’re not ready. So having that flexibility, that relationship with your cybersecurity partners, I think is going to be very crucial to success.

Ravi Krishnan [00:45:32]:
And putting together a governance plan upfront like we talked about, I think is going to be very important and at one level you almost want to connect the business to your security teams and you’re like facility at some point as the facilitator you can, your job is done. When you connect them and you get out of the way, let them work their problems out, right? When there’s a problem, that cannot be resolved, that’s when you step in but for the most part, when security questions things like is this PII data? Why is PII data needed to make your decisions? Can we just obfuscate this data? Can we tokenize it, put it on the cloud and you can just retrieve it on demand? We would like the business to say, no, I cannot function, I cannot make business decisions unless I have data in the clear in this environment. So having the business support your viewpoints when it comes to execution and things like that is very important to even success. So it all starts again to summarize upfront relationships, making sure that we are strong, we are on good talking terms and execute together. Giving them a seat at the table, at the outside, and then keeping the relationship smoothly functioning is very important for our success as CEOs.

Sanjog Aul [00:46:50]:
So Luis, when you look at like the way you explained your initiative is that you started it and you’re enjoying that journey, but at the same time you’re also supposed to in some form or fashion ensure that you embed security in the DNA of your whole data organization and also ensure that you put governance structure in place but when you’re yourself trying to build a house, if you will, how do you put such things into the DNA, especially when the structures keep changing, the needs keep changing?

Luis Fernando Lopez [00:47:23]:
Yeah, it’s a matter of foundations. No matter how big is the pressure, no matter how hot temperature, you need to have governance in place. If you stop having governance because you are trying to deliver faster, you’re going to crash at some point. Either you have to go back and redo the work or you just gotta plane crash it. The governance and the security thing, I really agree with Ravi here. They have to be on point since the beginning. When you do that, it may take a little while longer, but you make sure it happens, that things happen and it happened to me, not so long ago I’ve been with the year.

Luis Fernando Lopez [00:48:11]:
This was with another company. We created a solution that helped sales people go faster and making goals specific to each person in a sales floor and it was really cool. The thing is we never do with security and this was with a tool that you could have on your cell phone. So we delivered the app, we rolled out the app, we rolled out kind of the dashboards from the end of the project and we got stopped at the rollout. Even though the server was right, even though the information was protected, even though everything happened because we didn’t have security from the beginning.

Luis Fernando Lopez [00:48:53]:
We got stopped for two months until all the security issues were checked, until everything was revised and everyone saw the documents at the very detailed level and we missed opportunities of a winter sale. So it’s just if you don’t get security enrolled from the beginning, you’re just going to stomp out of the wall.

Sanjog Aul [00:49:16]:
So one last question for you Navi, organizational structure, culture, leadership, these are all softer sides but when you’re doing something like modernizing the data management where it’s part art, part science, what do you think should be the tenets of a conducive culture, a conducive organizational structure and effective leadership which will provide the necessary support for modernization data management?

Navi Grewal [00:49:46]:
That’s a good point that you bring up, right? Do I have the perfect answer? No, because we’re learning as we are moving along too, right. I think one of the most important things that I seen that works in most of the organization is a top down culture, right? When the CEO and when the all the leadership team is behind it and asking for it, and when your CEO said, hey, I am not going to look at your access, I’m not going to look at Excel, I am going to go into this place and that is where I’m going to get my answer. That helps. That gets the whole organization speaking the same language, speaking the same data and all that to me, that top down and then once it starts to. That’s when it starts and then everyone starts to look at those data sets and they start asking questions that hey, this data doesn’t look right. Can we go fix it? There is energy and then effort that goes into getting that started.

Navi Grewal [00:50:43]:
So they generally show that arrow where all the arrows are going different waves and lose energy when everyone is aligned going that one way. I think that has helped and data is not only a CIO or a CDIO or a Chief Digital Data Officer’s job, right? Data is everybody’s job. We want to make sure that we are a data driven company and a culture and that everybody has to be putting in their part. So to me actually the top down culture and, and the leaders getting behind that and making that a big deal. You have to make data a big deal. Any company that is not making data a priority are going to really hurt themselves and fall behind.

Sanjog Aul [00:51:26]:
Once again, thank you so much Navi, Ravi and Luis for sharing your insights about how organizations can take a step back, evaluate what they’re doing, understand the value of data, work with the management and as a result come up with a good strategy and execute it impeccably as much as possible, take care of the security and the governance along the way and of course keep improving their leadership jobs. All the things that we discussed towards this end. Holy grail of having a beautiful looking modernized data management approach which will help and serve the business well. So thanks so much again, lovely conversation, great insights.

Navi Grewal [00:52:08]:
Thank you.

Ravi Krishnan [00:52:09]:
Thank you.

Sanjog Aul [00:52:10]:
Have a good one and connect with us on social media. Subscribe to our podcast please. Once again, thank you for listening to ctn. This is your host, Sanjog, signing off till next week. Take care and God bless.