The CDO to CDO podcast is hosted by Chris Knerr, Chief Digital Officer of Syniti.
Subscribe Here: Listen to all the episodes in the Syniti Podcast Series
Chris Knerr (00:05):
Hello, and welcome to CDO Magazine interview series. I'm Chris Knerr, chief digital officer of Syniti, a world leader in enterprise data software. And we're partnering with Chief Data Officer Magazine, MIT CDOIQ and the International Society of Chief Data Officers to bring you this series of interviews with Global Thought Leaders and Data. Today, I have the great pleasure of talking with Adita Karkera, Deputy Chief State Data Officer for the State of Arkansas. Welcome Adita. Very nice to meet you and look forward to our conversation today.
Adita Karkera (00:36):
Likewise Chris, thank you for having me and this is a great initiative, so thank you for what you all are doing there.
Chris Knerr (00:42):
So as a starting point, could you give our audience a general overview of the scope and key services that the Arkansas Division of Information Systems provides? And then talk a little bit about how the Chief Data Officer role fits into that overall organization?
Adita Karkera (01:00):
Absolutely. The Division of Information Systems is the key technology provider for all of our state agencies and departments in the State of Arkansas. So we provide everything from hosting services, to network, to broadband, to support of the Arkansas Public School Network. So really we're the technology providers for all needs around all state agencies and departments within the State of Arkansas here. Now to come to the office of the CDO or the Chief Data Officer for Arkansas, the General Assembly passed legislation in 2017, that created the position of the Chief Data Officer for Arkansas. So Act 912 of 2017 created that position for us. And really the focus was or the need was to bring increased focus for creating data as a strategic asset for the State. I started in my role as the deputy chief data officer for Arkansas since 2017.
So I'm close to about three years, some in that role today. And in my role I work with different state agencies and departments to collaboratively create data policy, standards and procedures in the hope to improving data management and data governance around state agencies and departments. It's a very exciting role, very challenging, a great mix of technology and business needs, but with every challenging role comes a lot of responsibility dealing with government data and making sure that the right data is available and used to inform the right policy. It's a role of big responsibility, but a great role and I'm very excited that Arkansas is on it's journey to being more data-driven since the legislation was passed in 2017.
Chris Knerr (02:49):
That's amazing and we'll come back to this point a bit later, but I think that's very interesting and very enlightened that the State actually took the trouble and had the foresight to put in place a strategy in an organization. And it's terrific that you've been part of the whole journey from the very beginning. So maybe to bring that to life, are there a couple of key initiatives where having this organization and having your team, having a data strategy, better data provisioning has really made a difference that you could highlight it. And maybe if you could, because I liked what you said about how your role and the strategy encompasses both a set of technology outcomes and a set of business outcomes or what I would call client outcomes, maybe if you could make sure to touch on both aspects, that would be really interesting.
Adita Karkera (03:45):
So there has been several initiators that I would want to jump on and talk about just now as I'm going through my mind trying to think of one or two that I'd probably want to highlight just now. Probably one of the first things that I would mention is when the office was instated in 2017, the first thing that we started off with was what I like to call listening tours with different state agencies and departments. And what we did there was we went out to try to better understand where each department and each agency was in their data maturity journey, what were their data needs? What were their ping points? With that listening tour and an analysis of that, we came up with a publication for a gap analysis of state department needs. What was highlighted in network, two issues of data sharing, data governance and the lack of having a data catalog or an inventory of data assets. So different agencies and departments could know what each other had.
So we took that publication and moved on to our first project of cataloging or creating an inventory of data assets from different departments. That effort started in about late 2017 to early 2018. And today we have data inventory from over 15 departments collected together in an online portal. Together with not only the assets that they own, but also with the governing regulations around it, the data stewards, the critical information around those data assets are now listed together in an online portal. And that has been a tremendous win for the State. That inventory is available across the State. So different departments could see the information supply chains between departments, which is so critical when it comes to government. All of these datasets are so interlinked to each other, that we have to know how the information is flowing through these departments.
I think that has been a tremendous project that we have been able to get going. And like I said, we have 15 departments right now and it's still growing. We're still having more departments and agencies added to it. And it has been put into great use when we've implemented projects like our Statewide Longitudinal Data System. So the Statewide Longitudinal Data System is an effort to link administrative data from across education and workforce agencies and departments. It's absolutely critical for this to move forward, especially in view of the current crisis and pandemic that we are in to be able to provide insights into unemployment or education status, just provide the insights into the needs of our evolving economy today. And we've been able to leverage our data asset inventory or our catalog like we call it here, to see what datasets are available in which departments, how information is flowing and how those datasets need to be brought together to make more sense and be available to provide more informed decisions.
Chris Knerr (06:53):
So that's a great accomplishment and really interesting too. And I think for a lot of our audience who are in private industry and not in the public sector, that will be... I think it's particularly worthwhile to study what you've done and if you think about the scope and complexity, the stakeholders, the different agencies of a big state [IS 00:07:19], this is exactly the same thing. I see many large corporations struggling with and actually struggling to understand why it's important to do that foundational work that's going to actually drive the sharing and provide a solid foundation for any process re-engineering, outcome driven systems enabled or technology enabled programs. So you mentioned in that COVID and the COVID response, was the catalog important to how Arkansas responded to the COVID pandemic? And did your department play a role in leveraging that data to improve response outcomes?
Adita Karkera (08:09):
Yeah, that's a great question, Chris. I can't stop telling you how many different areas or policy initiators the catalog has come to be of value for, and definitely the pandemic. When the crisis hit us, there were so many different areas, obviously that were impacted and not just the health sector. And that's where we've been able to leverage it to say, where are the other datasets existing today that could be integrated, or at least what information is available that could be shared, which is relative to the pandemic. And again, it's not just the healthcare datasets that we're talking about. It could be information on unemployment. It could be information on childcare. When critical workforce is expected to come back to work for taking care of critical needs, they needed a place where are the child cares that are open and functioning while during the pandemic?
Where do you get all that information for? And in the many hats that the CTO wears, I think one of those hats is the role of a coordinator. And I think that's where we saw our role being leveraged to play the role of a coordinator to assist in identifying the datasets that are available and play that strategic coordinating role to bring together different departments. Another area which I've been fortunate enough to provide assistance in it, is Governor Hutchinson's COVID-19 Technical Advisory Board, which was created in May to get a group of technology leaders from across the state landscape and get them together to start evaluating the current technology solutions, look at contact tracing and also to provide recommendations on data-driven solutions to combat COVID-19. It is a great group of people and I'm honored and privileged to be serving with that group of people together to come up and start looking at what we need to do to move from a more reactive mode like we were when the pandemic hit, to more of a planned approach to take our economy to the next steps.
So the effort now is instead of responding directly to the crisis as it was in February or March, today being into a more planned approach and getting more insights into how COVID-19 is impacting different areas of our economy, whether it's unemployment, whether it's unemployment across industries, whether it's unemployment geographically, school data, what is the impact on our talent workforce? Is our gap between talent workforce available and the needs of the industry, is that widening? Insights like that are what the board is working on and data plays a very exciting role at this time.
Chris Knerr (11:09):
Well, that's amazing and congratulations on the work and then also on your personal role and your team's role on it. I think that's fascinating. And I've had this observation talking to many data and IT in business leaders that, organizations that were able to respond most effectively had the foresight to already have the foundational work in place, which I think is very much the story that you're telling here. So I imagine that if that Act hadn't been passed in 2017, your office hadn't been put in place that the scramble, and it was a scramble plan for everyone. The scrambled plan execution would have been a lot worse. And you might, as some organizations still are kind of be digging out still and not able to move from that diagnostic and reactive mode into a more predictive mode that really supports the recovery.
And I definitely share the view that I'm sure I'm preaching to the choir a bit, but for many data professionals it feels like we know how everything fits together, but all too often we're like the last person to be asked how to connect the dots between things. So I love stories where the opposite of that is true and that's very much what you're sharing. So that's really interesting. Maybe one more point, I like how your response illustrates the importance of not only a vertical data in the healthcare part, but horizontal data spanning across multiple domains and why making these foundational investments it's always going to be important, because when you need them, you really need them. And you may not always be able to think in advance like, "Oh, if we happen to have a global pandemic that was a healthcare crisis and an economic crisis and a human crisis all at the same time, what data would I need to respond to that effect of always?"
So I think that's extremely interesting and really illustrates a number of valuable points both in data strategy and executing a data strategy. Maybe I could shift gears a minute so, in talking about the challenges that you've been working through, you mentioned data sharing, you mentioned governance, you mentioned the data catalog. I noticed that you posted on Twitter recently that data quality remains among the chief challenges for CDOs. Can you elaborate a little bit on where you see specifically the challenges, are these technology challenges? Are they business challenges? Are they a mix?
Adita Karkera (14:04):
Great question. Again, Chris, I think it's a mix. I think data quality is first a challenge that I think is more process driven. What I see happening in organizations is the lack of two things. I think first, they need to identify what defines data quality for them. What are the requirements for data quality? Because they're going to differ from organization to organization, and maybe even have more than one definition of the same dataset for different consumers. So organizations often are failing to define what are their requirements for data quality. And then the second aspect of that would be to say, how are you measuring that data quality? . So, okay, now you have your requirement defined clearly. How are you going to continue to measure that in a manner that is comprehendible and is usable by everyone? And the more we talk about it, the more we discuss this as organizations individually and as organizations coming together from a statewide perspective, the more we are going to be able to make progress on it.
What we've done for that from initiatives on a statewide perspective is, today we have a data governance committee set up, a state data governance steering committee set up and that committee has representation from agency data officers from across different state departments. So we get together on a monthly basis and we bring together topics that are of concern from different departments and agencies. Now, whether it's related to data quality, whether it's related to changes happening in information supply chains in a particular department that's going to impact another department, we get together and we talk about those points every month. And those discussion points that are then taken back to the data stewards and the managers in each department. And that's how important topics like, quality governance, security and privacy are continuously addressed by the steering committee. And the idea is to create that movement, not just vertically or not just horizontally, this movement needs to happen in both the vertical chain and the horizontal chain to be actually effective.
Chris Knerr (16:19):
So I have the experience a lot. It's a question and an observation that I'll share with you. So I've had a lot of conversations in my life about data quality and very senior people often have the reaction. I don't understand what it is. It sounds expensive, who cares I've got better things to invest in. What I've found to be helpful in addressing that is to focus more on the outcomes that the data quality drives rather than the data quality intrinsically. Almost the way you might focus in manufacturing, having a good product at the end rather than measuring all the manufacturing process parameters, which manufacturing folks care about the people running the company and the customers they're like, "Is it a good quality product or not? Am I happy with the product?" Is that an issue that you've faced or I wonder if given that working in the public sector you have the opportunity to have a more foundational organizational investment that people more or less get why data quality is important. And I think that was about eight questions and six observations all rolled into one.
Adita Karkera (17:37):
I think, again, it goes back to the point of literacy around that topic. The more we talk about it, the more we get organizations to understand the value. What is the end result that you're looking for? And it doesn't... Like I said, a lot of organizations shy away from the idea, because of what costs are you trying to avoid? They're not clear what is the final gain from this. And a lot of times that that gain could be cost avoidance, it could be creating value, it could be re identifying brand image. I mean, there are so many non-tangible benefits of data quality which are probably the hardest to put down on a sheet. So those are the areas that the more we bring together these data leaders from different organizations and continue to champion and talk about these topics. That is how I hope that they're able to take those topics back to each state department or organization and champion those efforts within their organizations. And that's all.
Chris Knerr (18:48):
I think he said something similar to what I said in slightly different language, but so to go back, I mean, if I were explaining this to a public sector agency person who didn't understand, I might bring up what you shared about COVID and said like, "Look, if I can't understand who the people are, because their data's not properly maintained, then I can't provide the services that they need an effective way. And we're not able to get good outcomes for the people that we're trying to serve." And so I think that was your observation as well. I think that that education and the connection between the mechanics of the work if you will, and the data professional stuff connecting that back to the broader context is something that I think continues to be really important.
On related note, what's your sense of data sharing among different levels of government and to the extent that it's relevant private industry? So I'm thinking of, you have both an advanced maturing state agency, you have neighboring States, I'm sure there's relevant data you can share with them. You have the federal government, you have municipalities, which presumably are set up with different budget charters and so on. What's your sense in broad strokes of, are we happy or sad as a country, as it relates to that data sharing across different levels of the public sector?
Adita Karkera (20:23):
I think the public sector has made tremendous... We've seen tremendous growth in the public sector as far as the value of data goes in the recent years. Whether we start looking at it from the perspective of what the federal government is doing, that the Evidence Act or what the States are doing as far as creating more and more chief data offices around the States. And I think we're seeing the same movement happen at the city level. And the fact that more and more CDOs and more and more CDO offices are being set up, both at the state federal and city government. I think that itself is a huge indication of how seriously government is now treating data as an asset. The work that the federal government is doing around the Evidence Act is great.
We've tried to continue to look at it and follow some of those guidelines when we are creating policies for ourselves at the state level. Another thing that the state CDOs are doing there is a state chief data officer's network that has brought together to state CDOs from different States. And that network together collaborates and comes up with different policies and procedures that might be useful for different States to use, data sharing has definitely been one of the key topics that the network has constantly talked about. In fact, there was a recent publication from the network around data sharing guidelines. So I think just the fact that there's so much more emphasis and focus through this office of chief data officer that has been set across in public sector, that itself is a great testament that we're taking data as a strategic asset now.
Chris Knerr (22:11):
So I want to stick with this thought a little bit about the relationship between private industry and public sector. So if companies are interested in looking at, for example, the workforce in Arkansas and say, "I'm thinking of opening a new office or starting an enterprise in Arkansas." And they don't know anything about the data assets that you have available, are there workforce assets or economic sector assets that companies should be aware of and should be thinking about, I want to grow, I'm evaluating Arkansas perhaps versus different States. What kinds of things are available that would be useful commercially in that vein?
Adita Karkera (22:56):
That's a fantastic point that you bring up Chris. And interestingly I mentioned our State Longitudinal Data System earlier that we're working on. And as part of that effort, we're collaborating with the Economic Development Commission and the Department of Commerce, to identify the needs that is what you're pointing towards. When a new company is coming into Arkansas, how do they get to know what is our skilled workforce looking like? What are our college graduates looking like? Who is coming out not only today, but what is our workforce going to look like in the next five years to come? So that, that data can be used to guide industries that are looking to place in Arkansas.
And also to be used by our state departments and our department of commerce to say what industries can be attract to Arkansas based on not only the geographical location, but the workforce that's available. And more importantly, will be available in the coming years to come. So that's a project that is underway right now. We're very excited about it, because that's going to bring a lot of cost and operational efficiencies to some of the processes that are being done in a less automated manner today. So really looking forward to that, and maybe if we talk again in the next few months, I might be able to give you a better update of that.
Chris Knerr (24:17):
Super. And if we have to put in a plug for why data quality is important, now we could say, there's work underway, data quality is going to help us grow the state economy and everyone thinks that's important. So let me then shift gears one more time and talk a little bit about your career and your journey to this terrific leadership position that you have. So you started out, I understand in a very technical track on the database side, what was the transition for you from that technology track into the world of master data, operational data and how did you find it personally and professionally moving from a world where you were really worried mostly about systems landscapes and now your equally worried about systems landscapes and stakeholder landscapes?
Adita Karkera (25:14):
I have to say, I think it's just having the passion for data from the very beginning. So I know you point out that I started in a very technical role and I'd like to say that I still want to focus in that aspect of technicality in my job role today as well. But the fact that I've been with the state for about two decades, 20 years now, I have seen so much happen with the state data systems that we have managed over the years, seeing the problems that we fixed, seeing the information that is flowing between departments and the challenges that we've addressed and helped figure out over the years, got me to start thinking, I want to know where is the data coming from? Where is it going? How many copies of this data exists, not only in my organization, but around state department? How many times are we collecting from the same data again and again?
There was so many opportunities for improvement. As every day of the last 20 years has been a learning experience for me to learn more about, what can I do with the data that we have at hand today? What is the department using it for today? And more importantly, what can it be used for tomorrow? What can it be used for tomorrow if we integrated it with other datasets? I think it's the curiosity and the passion for trying do more with that data has transitioned me from day one of my journey here at DIS to where I am today. And again like I said, it's a learning experience every day.
Chris Knerr (26:56):
That's amazing. Well, and I just... This is the first time we've met, but I feel like you do have that passion and that willingness to educate other people about why it's important. And in my experience that's really helpful to graduating into these senior leadership roles in the data world. Well, maybe in wrapping up, any major predictions in the data world on say a five to 10 year horizon, or final thoughts or messages that you'd like to leave the audience with?
Adita Karkera (27:33):
As far as predictions, I think two things that I like to focus on are going to be valuable, cloud computing. I mean, we've seen cloud has been a game changer in the last several years. In the coming few years I see a bigger shift in cloud data management, but I also feel as DNA leaders, I think we need to prioritize, we need to decide what we want to move over to the cloud. Cloud has its own potential benefits, but I think we need to always decide the right service for the right use case. So we're able to use cloud to bring the cost and operational efficiencies that it gets. So as DNA leaders, I think we need to continue to evaluate and prioritize them. And I see that being more of a hybrid shift, being more of a cloud right approach than just a cloud first approach.
The second thing that I think is going to be valuable is responsible AI. AI is here to stay, but I think with that, even at the current pandemic times, we know that AI and machine learning has played a huge role in providing insights and predictions around the virus. But together with all of these potential benefits that AI has provided to various industries comes a great deal of risk. And I feel that as we're creating more AI frameworks, there is a need to wrap data ethics and privacy in these AI frameworks.
We as DNA practitioners and leaders, I think we need to be careful about being able to explain these AI frameworks. Everything needs to be governed with the right ethics and privacy while creating potential benefits, because yes, AI is being used to drive some very critical predictions today and those predictions might be impacting the lives of some citizens. So let's do our part in being responsible as we apply some of this. Wrap up thoughts, I guess, to me learning is important. I just say, "Keep learning, it doesn't matter whether you're day one on your job or you are 50 years into your job, just keep learning." I think that's one piece that I always like live it.
Chris Knerr (29:58):
Yeah. That's perfect. Just to interject, I think those are great predictions and I agree in particular on algorithmic transparency, responsible AI, I think that's going to be really critical and maybe just to tie it back to one point in our conversation, I think for our audience, no one's going to do this for us. No one is going to do it for us. So people who understand this and who are in leadership roles, we have to be the evangelist for data ethics and AI responsibility, because we understand it and understand that both outcome levels and then also technical levels how, for example, bias can creep into AI models and there's been a lot of excellent and there continues to be a lot of emerging research on the propensity of those models to bias of the folks who write the algorithms, bias and datasets and many other aspects of this.
And I think you're right that this is dawning on everyone in the world and on a five, 10 year horizon, people are not going to accept black box results, especially when they come to consequential things like, state service provisioning, or approving of credit loan applications, or healthcare or many other things. So I think that's a great observation. Super well, this was an awesome conversation, Adita, thank you so much for joining me today. I really enjoyed meeting you, for our audience we have multiple other interviews at cdomagazine.tech and thank you again. I hope everyone has a terrific day.
Adita Karkera (31:43):
Appreciate it. It's been a pleasure. Thank you.
Chris Knerr (31:43):
About the AuthorFollow on Linkedin More Content by Chris Knerr