Data-Sharing & Data-Driven Decision-Making Task Force
Video
Transcript
4 documents
Machine transcript
May contain errors. Verify important quotations against the official video.
About transcript accuracy
- Source
- Whisper
- Model
- ggml-large-v3-turbo.bin RTX5060
- Processing date
- October 8, 2026
Unknown speaker
0:00
Thank you.
Thank you.
Thank you.
Thank you.
Thank you.
Thank you.
Thank you.
Thank you.
Thank you.
Thank you.
Thank you.
Thank you.
Thank you.
Thank you.
Thank you.
Thank you.
Thank you.
Thank you.
Thank you.
Thank you.
Representative Stephen Meeks
Unverified
10:00
All right, we're going
to go ahead and kick things off. the official gavel here. Because today is the 18th anniversary of 9-11, I believe it's appropriate
for us to start today's meeting with just a moment of silence and remember it's all the Americans who lost their life on that day. So please join me. Thank you. I appreciate everyone being here this afternoon as we continue our work looking at the future of data here in the state of Arkansas and I'm going to begin the meeting by I guess turning it over to my co-chair see she's got any comments to get started this morning. Oh
Senator Jane English
Unverified
11:10
thank you I once again I think this
is probably in my mind one of the most important things we can be doing in state government to move us all forward to have a better idea of using our data how we can use it what we could be doing with that um so this is important and i
Representative Stephen Meeks
Unverified
11:31
appreciate everybody being here thank you totally agree uh so with that uh we need uh approval of the minutes from the august 14th meeting you should have a copy of those uh i got a motion second all in favor say aye
any opposed all right we got those approved and then we'll get right into the heart of our meeting this morning. Is Ms. Jones, Ms. Jessica here or y'all? You're filling for us. Okay. So if you would go ahead and have a seat at the table, just I think most of us know who you are, but please introduce yourself to the committee and then you're recognized to
Speaker 14
12:10
proceed. Do you have anything up here? No. Good afternoon, chair and members
Aditha Karkara
Unverified
12:13
of the task force. My name is Aditha Karkara.
I think so. It's green. Is that better? My name is Aditha Karkara, and I'm serving as the Deputy State Chief Data Officer with the Department of Transformation and Shared Services, Division of Information Systems. Director Jones is unable to attend today so my team and I will be representing her on her behalf. In last month's task force meeting we heard from the CDOs of Indiana and North Carolina on their state's data strategies. We ran out of time we were supposed to be talking about the data
catalog and we were supposed to be talking about the cost and feasibility study so those are some of the topics that are lined up for today's meeting. There should be a handout available for the task force that we can walk through.
Speaker 25
13:09
yes everybody has a copy of this correct okay
Aditha Karkara
Unverified
13:14
the topic of the state data catalog is probably not a
new one for the task force but I'm just going to take a minute to give a brief overview of what it is just in case someone has not heard about this before. It was mandated by Act 912 that we have to start cataloging the data that is hosted by all the state agencies and as part of that effort the Office of the Chief Data Officer started work on cataloging the high-impact data assets that are stored in different state agencies. The output of that effort today is in the form of an online data
portal. The link to that is part of the handout. Each agency today now has access not only to continue to view and update the inventory for their state agency, but they are also able to view the data assets that have been collected by all other state agencies. That is a change that I want to highlight from my last presentation because the last time we talked, each agency was only able to see and update their data. Today we have opened it up so that all
agencies that are participating in their effort in this effort are now able to view data from other state agencies. This is just another step to broaden the scope of this effort and allow more transparency between the state agencies that are participating in this effort. Each agency's data officer now has permissions to continue to update that inventory and enrich the data that they're hosting with their agency. As we continued on with this effort, we realized that there are issues with the data quality of the inventory that has
been collected. So this is a process that is going to continue to grow and evolve as we move forward and continue to add different agencies. We understand that the current agency data officer from each agency today is just an existing state employee that is functioning as an agency data officer. So So they're trying to do the best they can in their capacity to enrich their data. The plan moving forward is to try to continue to identify agency data officers in new agencies that are joining this effort.
As part of transformation, there are several other agencies in different cabinets that have come up forward and said they want to be involved with this effort, which is great news. So we're continuing the effort to work with those agencies and identify data officers and hopefully create a Governance Steering Committee which has been proposed and approved by the Data and Transparency Panel in one of our previous meetings. The scope and the vision for the Governance Steering Committee is that it's going to have representatives from different agencies' data officers.
They're going to meet periodically and define operational standards. They're going to review the state data assets that have been collected as part of this data catalog to continue to identify high-value data assets that could potentially feed into an open data portal. The steering committee will also continue to assess business use cases that may come forward from different agencies, create a state data governance plan, and have standardized data management practices set for the state. That was a quick update on the data catalog effort that's been underway I'm open for any questions okay I've got a
Representative Stephen Meeks
Unverified
16:34
couple of questions but before I jump in representative love will let you start sir
Representative Fredrick J. Love
Unverified
16:42
thank you mr. chair I'm I'm just curious on now what kind of data are we cataloging I mean I guess you just said for agency like DHS what kind of data are they cataloging currently
Aditha Karkara
Unverified
16:59
let me just clarify that when I'm talking about data it's not actual data but these are data assets for that agency so it's not actual data but maybe a data system that
let's say DHS or DIS or another agency is hosting so it's an inventory of the data systems that are being utilized by that agency the kind of data sets that are included have really been decided by the agency
Speaker 14
17:28
itself the recommendation from the office was try to inventory data assets that are high impact
Representative Fredrick J. Love
Unverified
17:33
for them okay so data assets are available okay so let me ask you as a follow-up I know earlier we're talking
about data sharing agreements. Where are we on that
Aditha Karkara
Unverified
17:47
effort? Are you referring to the multi-agency data sharing efforts that we've talked about there? I know that is still a constant effort that's being looked at. There are several other inter-agency data sharing agreements that are being utilized today, but as of having a statewide multi-agency
Speaker 28
18:04
data effort, that is still just work in progress. Okay. Senator?
Senator Jane English
Unverified
18:16
So let's go back here where it's the
next steps. So this governance
steering committee, how different is that going to be than the transparency panel? Madam Chair, the
Aditha Karkara
Unverified
18:29
data and transparency panel is comprised of executive leadership from state agencies and some representatives from the private sector, it does not comprise of data stewards or people who are directly dealing with data in and out. So while it's great for making decisions at an executive level, they do not have the time
to meet on a as-needed and periodic basis to address very minute issues with data. So the vision is that the steering committee would be comprised of people who are dealing in and out with the data needs of their agencies so they can talk to
Representative Stephen Meeks
Unverified
19:19
that as needed. Okay. Okay. So my follow-up question to that is,
do you see the governance steering committee as a standalone committee or are they going to be kind of a subcommittee to the data panel so that they're ultimately answering to the data panel? Is that the correct vision? What's the hierarchy between these two committees,
Aditha Karkara
Unverified
19:42
I guess, maybe the better? Yes, we consider that to be a sub-working group or so of the data and transparency panel. Eventually, it's the data and transparency panel that's making decisions at a higher level. We just needed data stewards and data officers at a lower level who are able to come together, define their needs, and be able to help us create a state governance plan.
Representative Stephen Meeks
Unverified
20:07
So the people who are doing this day in and day out are
the ones who are, yeah, sounds like there's some wisdom there. The other question that I had is, and you may have already answered this, but it seems like there's been real good participation in this. So the agencies that have been involved have been actively making their catalogs available and have proactively been engaged in this. Would that be correct? Or are you seeing any agencies that are hesitant, providing pushback?
What are you seeing as far as that goes? I
Aditha Karkara
Unverified
20:42
think we've seen a huge culture change in the last two years since we started this effort. it it was something new for the state when we first started this so it took us time to work with each agency and help them walk through the process it was a challenge in the beginning just to get the cataloging done but but today I see that several state agencies are now understanding the value of that effort
they're starting to see why we need to catalog this and how it could possibly come together to help them make better decisions so so I'd say we've made huge progress in terms of a culture change for the state as a whole and people are and the agencies and their leadership are wanting to come together and pull in to this catalog to enrich what's available for the state to use. So I think I don't see any barriers or issues I think we continue to propagate that culture and get us to a more mature data catalog. Awesome, that's that's
Representative Stephen Meeks
Unverified
21:41
encouraging to hear because I know it was not that way when we first started.
Speaker 56
21:49
Representative Brown, I think you were next, ma'am. Thank you, Mr. Chair. And I know that you have answered this, but I just want to be perfectly clear in my mind. When you talk about a data asset, you're talking about a category of information, like name, that an agency would have somebody's name, they'd have their address,
they'd have different categories of information. That's an asset, and that's what the catalog contains, like this agency has this, this, this, and this information, but if I am a member of that agency, I can't go in, I can look and see what another agency has in their catalog, but I can't actually look at the data, like any of the real information there yet. Is that correct?
Speaker 28
22:43
Yes and no. So a data asset, if I may give an example to maybe further elaborate on that, a data
Aditha Karkara
Unverified
22:51
asset could potentially be ACEs. I think everybody here is familiar with what ACEs is. ACEs are a financial system for the state. That is a data asset. We have not gotten to the level of detail to say what are the data attributes, what are the pieces of information included in ACEs
Speaker 58
23:13
yet. so my question about name address whatever that's the pieces that you're referring to
Aditha Karkara
Unverified
23:19
that's correct representative okay we have not gotten to the point to start enlisting what are the names addresses tax ids whatever other pieces of information included in that data asset we do not have that enlisted as of today so what you have now is really a really broad view of what each agency has that is correct representative it's only a very high level inventory of what are the systems that your agency may be hosted. And we think that that's really high and broad, but it was surprising to see that we did not have
that information when we started with. Okay, thank you.
Speaker 62
24:00
Representative Love, you're up against her. I do. I
Representative Fredrick J. Love
Unverified
24:03
have one more question. Before, I think it was either the last meeting or the
meeting before, we talked about business case and we you know there was no kind of a definition of a business case moving forward are we looking at are we looking at how we're going to define what a business case would be in order to tap into the database or to to pull information out some some of those discussions
Speaker 14
24:28
are happening at the data and transparency panel meetings on you know what use cases mean how do
Aditha Karkara
Unverified
24:34
identify what is a use case. I'm hoping that we can continue to further elaborate what a use case is through our governance steering committee. That is one of the that is one of the topics that I hope that the steering committee is going to be addressing. How do we assess business use cases and how do we define what is a business use case? What is the ROI from a business use case, a return
of investment from it, before we decide that yes it's worthwhile pursuing that use case okay all right thank
Representative Fredrick J. Love
Unverified
25:04
you it's good to know that you are looking at that because that was one of the questions that I really I really had on the top of my mind it's like all right how do we define this and really you know so thank you yes sir so could
Senator Jane English
Unverified
25:18
you just give us an example of what a business case might look like for everybody kind of so they understand a little bit more well what a business case look like
Speaker 14
25:29
madam chair may I point out probably a use case for recidivism so we have previously
Aditha Karkara
Unverified
25:35
presented to the task force some work that we had done on how do we reduce recidivism that that was a use case it's Arkansas has the fastest growing prison population in the country we we considered that to be a big issue for the state and for our governor that was presented to us from our correctional agencies and said, how could we address the problem of recidivism for the state?
That was a great use case because we could use our data, use our technology to showcase what we can do with data to solve the problem of recidivism going forward. Does that help? Yeah. And
Representative Stephen Meeks
Unverified
26:21
since that was so successful, is that going to go to a live status and be used on a
Aditha Karkara
Unverified
26:31
we at on that use? Representative, we went ahead and did all the analysis
regarding that proof of concept and passed it on to leadership to make a decision on that. So I do not have a definite answer on when we would be able to operationalize that. If you can introduce yourself, if
Representative Stephen Meeks
Unverified
26:48
you have a little extra, you want to throw one to us.
Speaker 72
26:52
Sure, Jennifer Davis. I'm the chief privacy officer. You're going to have to excuse me. I do have a bit of a cold, so if I'm coughing I don't think it's typhoid so I think we're good. So I just wanted to clarify that you know that was one specific case that we are really looking at and analyzing you know
how we can use this from a statewide perspective not a DIS perspective but what is the needs of the data transparency panel what are the needs of the state that was one specific instance and we're not anywhere necessarily looking at a specific solution at this point it is more analyzing what we can do as a panel to kind of remove those barriers from a legal standpoint with data sharing, which that was a question that was brought up, you know, try to put together a legal framework for how agencies or departments, excuse me, force of habit, you know, how those
departments, you know, foster care children at the age of 18, you know, age out. I'm not sure about you guys, but I certainly was not prepared to be on my own at 18. I'm not sure that I'm prepared to be on my own now. Let's just put that out there. But at 18, what needs do those children have that maybe aren't getting met from an education standpoint, from a workforce standpoint? That may be something that DHS might want to look at, or they might want to pull in ADE. They might want to pull in other workforce. They would collectively come together and say
that they have this need, and what can they do as a whole? It's not just any one person. It's really what can the state and those individual departments come together to say what can we do about this and how the data transparency panel can put those the right people in the right places to be able to put that together to make that decision does that help all right uh senator bond
Speaker 55
28:35
i think you're next sir well i think i may
Senator Will Bond
Unverified
28:38
have jumped ahead to the next agenda item so i can wait which just the report right I guess yeah so I can wait when we get to the report okay representative
Representative Fredrick J. Love
Unverified
28:49
love you're back up sir thank you mr. chair so that that was perfect I needed that all right so but when we began to talk about use cases we we said business use cases now we're moving to use cases so are we are we talking about internally as far as the internal state government structure Or are we talking about private businesses being able to tap into and use this information as well? Because I think that, I mean, that's kind of a, to me, when you talk about the internal stakeholders using the information, that's one thing.
The legislature, the governor, some of the governor's staff, you know, that's one thing. But when you start talking about private businesses using this information, it's something totally different. So what are we talking about? Are we talking about internally or are
Speaker 72
29:46
we talking about? Right. And we are talking about internally about how, and I say business case, I say use case. I mean it, you know, interchangeably. What need do the members of our data transparency panel have and how can they come together to make that, you know,
example that use case that business case i use that interchangeably in order to solve those needs to further the state of arkansas so that is i am referring to internally and you know as had been mentioned you know the idea of use case does need to be defined as well as you know the governance structure by the data transparency panel about you know whose use case comes first the priority those kind of things that are currently under works in the data transparency panel by the departments.
Representative Stephen Meeks
Unverified
30:30
Okay. All right. Thank you. And in my mind, I see it as ultimately as both to allow
for some of the data that's, you know, properly been cleansed and stuff to be available for use for private or for even individuals. You know, as an individual citizen, all of us have constituents who want to know, well, why is that tax money for highways being spent on sidewalks and stuff like that and wouldn't it be nice if there was a website and they could see where highway money where the revenues come from and where it goes and I think that you know can maybe even go a long way to help our constituents become more knowledgeable about how state government works
and potentially where there are businesses where we interact with that can either be a benefit to the state that there might be some potential you know researchers people who can you know at our universities can take this data, run research projects on it, and say, hey, here's a place where the state can do something better. So I see kind of an
Representative Fredrick J. Love
Unverified
31:32
all-of-the-above vision. Well, it was just one thing that I was concerned about, Mr. Chair, when we started talking about the privacy. And so that's why I really wanted the differentiation, because when you begin to
talk about the state of Arkansas using the data, it's one thing. But when we start talking about private entities using the data when we start talking about privacy of individuals that may be another thing sure so so that's that's why i wanted to to kind of find out if it was a you know the differentiation between them because i guess if we're selling if we do sell the information and we have the well she she looked at me so i guess we won't
Speaker 85
32:10
be selling the information so never mind you
Speaker 72
32:13
know there are some departments that already do publish some data i'm not sure
how familiar you are with the my school info from the department of education if you're not familiar with the tool you should really go out there and look at it it's really neat that you can go pull all the schools in the in the state or the districts that are maybe in your congressional district or in a county or look at statewide and they have an entire set of school and student related data that is completely cleansed that is available for the public that any member can go out there and look so a lot of departments are already publishing a lot of their own completely de-identified anonymized data sets and so they're already doing that in any agency at this
point is welcome to do that but if you haven't looked at my school info I would encourage you to do so
Representative Fredrick J. Love
Unverified
32:59
because it really is neat so is this information like realtors used to I like to like when they identified the school school districts
and different things like that is that I mean I'm just asking i would imagine
Speaker 72
33:09
it's something like that but i i know that when i had worked at department of education you know a lot of times parents would go out there and look at it because they were potentially going to buy a house in a county or in an area and they wanted to compare school so a lot of like i say departments are already publishing a lot of their own data um
that it again is completely you know the privacy is taken
Representative Stephen Meeks
Unverified
33:35
into consideration for that so that's already being done another example the geographic information office makes a lot of data available If you use, like, Google
Maps or Waze, a lot of that information originally comes from our Geographic Information Office. And to answer your question about privacy, obviously that's a chief concern, which that's
why we have her. So she's our privacy officer, and so obviously we want to make sure that that is taken into account.
Senator Jane English
Unverified
34:02
Jennifer, one day when we were talking, we were talking about the recidivism thing. And one of the things that you brought up was just where we were talking about health care and mental health availability and all that. And I think one of the things that we talked about was at the same time this was going on, DHS was looking over here at some health care centers or mental health care centers. There was no opportunity for, if the people had been talking to each other and the data had been shared,
probably could have made better use of those dollars and they could have been more, it could have been used for the recidivism as well as for the, I mean, there could have been a joint effort
Speaker 77
34:47
there rather than people trying to do everything on their
Speaker 56
34:51
own. Representative Brown. Thank you. This may be a question for the chairs, but I'm just curious, is
Speaker 58
34:57
there any chance that we could ever have any interaction with this data transparency panel and find out how they're
doing, or is our interaction strictly with our DIS department? Yeah, so the data
Speaker 92
35:09
panel, all the meetings, they're public and open, and
Representative Stephen Meeks
Unverified
35:13
I think I've been to all of them except for, I don't know, one or two, maybe. And so, yeah, and she's actually the chair of the committee. So, yes, and that was one of the recommendations I was going to make is if opportunity allows, I would encourage you, if you've not, to sit in on the data panel so you can kind of see the discussions that they are having.
So go to their meeting, not bring them to our meeting? Well, I mean, we can have them come to our meeting and discuss what they're doing. But essentially, I mean, she's the chair, and so she can tell us what they're doing. But to actually go and sit in on their meeting, I think, is more informative because you can see who the members of the meeting are, what concerns they have, and whatnot.
Speaker 56
36:08
I don't want to bog us down on that, but do we know, is there where those meetings published? scheduled. Yeah, the meetings
Speaker 95
36:15
are held here at the Capitol. Do you have
Aditha Karkara
Unverified
36:18
your next one scheduled? Yes, I believe the next one is November 4th, and I'll be happy to share that
Speaker 14
36:25
calendar invitation with Mr. Gilliam, and he can maybe share it with
Senator Jane English
Unverified
36:29
the task force. Thank you. I appreciate that. We've invited them to come to our task force meetings. Blake sends out an invitation and a notification of when we're having our task force meeting. So yes, well, you know, we'd love to have
them come. In fact, we've got one of their members is here and is going to make a presentation a little bit. But we even need people, I'd like to have auditing come and give us some thoughts and ideas about because we need some of
Representative Stephen Meeks
Unverified
36:59
these other people to talk to us a little bit as well. Very good question. Because we need to be interacting with them. So definitely. Any other questions before we move to the next piece?
Speaker 14
37:12
Hearing none. Okay. Thank you, Chair. As
Aditha Karkara
Unverified
37:18
the Office of the CDO realizes that data is a strategic asset for the state, and now with the data catalog effort and through several conversations with different agencies and departments and cabinets, we realize that they are encouraged to make the quality of data better for the state. Since we are moving more towards being data-driven, we realize that there are data quality issues across the state data assets.
With that effort and with that need in mind, the office of the CDO coordinated a proof of concept with a third party to see how we could continue to enrich the data that is hosted with the different state departments and also see how we can possibly create a unique identifier to link data sets between different agencies and departments. With that two-fold goal, the Office of the CDO collaborated something with DHS and the correctional departments to see how we can do a better match of the incarcerated population
who are maybe getting Medicaid benefits, or rather to find out that the incarcerated population is not getting the Medicaid benefits. I would like to call Mr. Jonathan Askins, who is serving as the vice chair for the Data transparency panel and he's also senior director with axiom to come and present the results of the proof of concept and give us some
Speaker 41
38:37
more details on that if you would sir when you get settled in
Representative Stephen Meeks
Unverified
38:43
just introduce yourself again to the committee and then you recognize
Speaker 105
38:51
make sure you're yeah that work okay thanks i'm jonathan askins and i am one of the directors for government services at Axiom Corporation. Representative Meeks, I want to thank you very much for having the moment of silence. Not many people know that within three days of 9-11, 2001, Axiom pretty much opened its doors to law enforcement from all around the world.
Basically, we hosted, at one time, we hosted 20 different agents who were working side by side with our analysts for about three and a half, maybe four months, because the need was very strong at that point to find the bad guys. And I can proudly say that Axiom never presented a bill to the government. It cost us a few million dollars to get that done. So it was a proud moment for us, a proud moment for me as I watched
that team put that together. And I think we've been told that we made some major progress. So you can be proud that you've got a company just 20 miles up the road who greatly contributed to 9-11 and the efforts after that. So I am coming to you today. I put my Axiom hat on, I guess. let me take that off. I'm vice chair now of the data transparency panel. And so what we did
is we wanted to have a proof of concept to where could we actually identify individuals at a good enough rate? Because if you're going to do anything from sharing and transparency, the identity and the ability to, with great certainty, to say that you have that individual, you know that individual, is critical, because especially when you're dealing with instances of marketing,
it's not as critical, but when you're dealing with operational systems and you're dealing with systems that might be looking for fraud or for those kinds of behaviors, having the identity right is is pretty important so we wanted to make sure that we could do that so we we intentionally picked dhs and corrections because we felt like that their data would probably no disrespect to them but their data would probably be some of the dirtiest that we could find in
terms of lack of elements or in terms of folks providing false information could we overcome those obstacles we were able to do that in the correction area we were not able to get to the DHS data for a variety of HIPAA reasons so we just decided you know let's I mean we could if we really were to move forward in an operational situation we could probably overcome those but when you're talking about a free POC nobody really wants to bang against the law to try to
get that done. So we moved on and just said, let's just go with corrections. We were able to identify about 77% of the time with a first pass. So you were able to kind of look at, you had a lot of different things. You had name inversions in there, you had misspellings, you had inversions of the SSN, you had no addresses being provided, you had false addresses being provided. So you take that first pass at a risk repository, and we were able to get about 77% matches, which is pretty good.
But then we began to add some analytics on top of that, where if we took, let's just say one of the second passes we might take is if we didn't have an address of an individual, yet we showed that we could see relatives of that individual through somehow they've used the same SSN or something, we could take the address of one of the relatives and apply that to the individual, and then sometimes we did get a match. So you apply a lot of different business rules on top
of your first run, and our match rate improved to greater than 90 percent. So when you're dealing with the corrections data, I felt like 90 percent was a pretty good job. So we felt good about that. Some of the things, I mean, I'm not going to bore you by going through the entire report, But some of the things that jumped out at us, and I think they jumped out at Secretary Kelly as well, is 118 records. This is our active prison system, our active prison population.
118 records showed that they were using Social Security numbers that were identified as deceased. 117 of the records were never issued by the social security association so essentially what you've got is someone just creating a social security number and they've used that for a long time and we were able to identify that of the population of 15,000 roughly 1,550 had social security numbers associated with them. I'm sorry, were using an SSN that had more
than four people associated with that SSN or in use of that. That typically says fraudulent activity. I'm not saying it definitely says that, but it typically says that. It certainly is a red flag to start looking at. We did find six records that had high risk indicators. I forget what all the high risk indicators were, but we had six records that were flagged for being really bad guys. So we did pass this information along to Secretary Kelly. And, you know, so maybe she can
get, maybe they will get some benefit certainly out of the POC, but the critical thing that we wanted to try to understand here is can we get solid identities, can we assign at that point some sort of identity that could, that would remain persistent with them, and And we feel with a greater than 90% chance that we can do that.
So essentially what's next out of this is, and Representative Brown, I'm sorry, Love, I'm glad you brought up the use cases. So what we're trying to do now is to develop use cases that would work for, now we want to bring in a second agency and run a similar process in a second agency and establish those IDs. And then after you do that, we can start to look at the overlap. So is there a case for
where an overlap might save us some money? And I'll give you an example that we're looking at right now. Let's just say we looked at labor and workforce services, okay? So if they're getting unemployment benefits, yet they're in prison, maybe that's something that we should be looking at. So yet that connection hasn't been made because the data over here says one thing, the data over here says another. But if you run it through a very similar system where you can
assign that unique ID, then you can make that match. And that's what we're attempting to do right now. We're looking at three different use cases. And from my perspective, just because I'm kind of a business-minded guy, I like to look at things, what's it going to mean from a money standpoint? We looked at one of the things that brought our attention to this was Elizabeth Smith's work with Medicaid, understanding the fact that the prison system, we were still paying
those Medicaid premiums when we shouldn't, the state of Arkansas shouldn't have been doing that at that point, that saved us, I heard $2 million. Don't quote me on that, but I heard some significant dollars. And so when you're looking at things like that, I think that's one of the things that we want to do is look at, okay, so what's a use case that makes sense? Does it make sense from a business standpoint. Is it going to save us some money? And then move forward with
that. But then the second thing that I think we also should look at is a lot of times people forget just the sheer fact that information is being shared allows the government to fulfill its mission more simply and more efficiently. That sometimes gets lost in all the money that we can save in fraud, a lot of times people forget that we have a mission, and that is to deliver goods and services and to deliver those efficiently. So it's a little bit harder to
quantify, but we want to be able to try to put some quantification around that to say, how can we do that? So fraud and waste and abuse are always going to get the lion's share of the attention because there's actual dollars associated with that. But the other use cases that we're looking at is, how can this make us more efficient? And could we put, you know, could we quantify some dollars to that efficiency? So that's really kind of where we are with the POC at this point.
I, you know, I invite any of, as vice chair of the transparency panel, I'm an appointee of Governor Hutchinson. So I invite any of you, if you would like to come and visit our panel, we'll make sure that we roll out the red
Representative Stephen Meeks
Unverified
49:06
carpet for you. Awesome. We appreciate that. And obviously this first simple
use case was very eye-opening,
showing the potential of what this can do for our state. So I guess my question is, we're starting to see some good proof of concepts with these test cases. What are the next steps to making these operational? or how soon can we expect to see, you know, what are the barriers that were?
Speaker 115
49:36
Well, do you want me to answer that or
Speaker 105
49:39
do you want to answer that? I can, okay. So I think there's a feasibility cost study with a recommendation in terms of a budget for a data warehouse.
The transparency panel interviewed other states. We interviewed other systems that were similar, and I think we've put together a budget that has, you know, a recommended annual budget for that. I would think that that would begin to be the next step from that standpoint is to begin to allocate the dollars for that, go through a good thorough RFP process, and find someone who can deliver the types of services that are required.
Aditha Karkara
Unverified
50:32
been created or worked on for certain stakeholders so at the end of each use case what we do is for that proof of concept go back and present the results to stakeholders and and present the options available for them to pursue it or help us move forward with it. So that's what we did with this proof of concept with Medicaid matched
with incarcerated population, and that's what we did with the other use cases. We compiled the results and presented it back to the stakeholders and said, okay, this is what will be
Speaker 119
51:05
needed to operationalize that. And it sometimes differs use case to use cases
Representative Stephen Meeks
Unverified
51:10
on what the next steps might be. And I think that's one of the challenges that we're going to have to look at as lawmakers is the, as we look at budgeting for the data warehouse, do we do it as a entity in and of itself
or do we somehow make each agency pay a piece of that? I mean, those are things that we're going to have to all figure out. And I will say that the last comment I'll make and then pass it over to my colleagues here is I like the idea of the efficiency and simplicity because to me, I'm looking forward to the day when the citizens of the state start to see the benefit. I think early on it'll be all of us in government, but when the citizens of the state start saying that, hey, their lives are a little bit easier and a little bit simpler because of this project,
then I think we'll, you know, we'll have done a good thing. Senator Bond, you're up first, sir.
Senator Will Bond
Unverified
52:04
Well, and I still may be jumping ahead to the report, but they ask you a question about what I wanted to hear from you with your experience out in the private sector. So on the the report that we were given this I guess the first handout we've proposed future Arkansas state data hub and so I guess that's the I think you were sort of endorsing that model but my question which may
be very basic is who's in the middle there is that and that may be representative Meeks what you just asked who is that a state agency is that DIS and is that where all the data is or is there an RFP so I mean how does that how do we I'm having a hard time thinking through how this committee helps us get to this
Speaker 105
52:51
the decisions we have to make it I'll I will address in terms of where the data is and and and I guess the concept behind this the concept is in the middle is
actually if it's a Rosetta stone between the agencies if you will so it allows the agencies to act autonomously and keep their data and actually keep an identifier that is unique and specific to them, okay? And you do that for privacy reasons and for breach reasons because if someone happens to breach the Game and Fish database, okay, and they get the unique identifiers, it won't do them any good outside of Game and Fish.
They couldn't go to DF&A, if you will, and say, okay, I've got the identifier, let's pull this person's identity, it won't work because the Rosetta Stone is in the middle, okay? So the master ID, if you will, resides in the middle. I have no idea if that's a DIS function, probably should be, but the master resides in the middle. Autonomous links or identities are outside and in the circles.
And you're not actually, so DF&A would certainly not be sending all of its data to the middle. DF&A would be sending some identifying information, perhaps, to the middle to get that unique ID, and then the unique IDs would be passed. And then ideally, how it works in the commercial world right now is, I mean, so if you took this, and this is a bank, if you will, okay, so the master ID sits in the middle, this is mortgage, this is checking, this is marketing, this is, all of them have their own autonomous IDs, and they have a Rosetta Stone in the middle.
And what happens is instead of passing large chunks of PII, you're now passing just a unique identifier. And that unique identifier is usually one that doesn't have any logic to it. It doesn't have any math behind it. It's just an ID. So if you do happen to breach that transmission, there's really no harm there. So from a privacy standpoint, that makes sense to just start passing IDs back and forth between the agencies rather than data being transferred.
Senator Will Bond
Unverified
55:23
Does that answer your question? I think it does, and I agree with my very limited understanding of this. I think from a security standpoint, my understanding is that's the best way to go, right? and then you do envision I guess a state agency which is what I have thought maybe everybody's way ahead of me on that would be the Rosetta Stone in the middle so then I may be jumping ahead and reading the next report is it 8 million or sometimes to me it reads like you need
4 million and then 8 million meaning 12
million sometimes it reads like you only need 8 million so am I jumping ahead? Senator I
Aditha Karkara
Unverified
56:02
think that's a great segue way to move on to our next topic. And I will like to call
Speaker 14
56:08
on Robert Magoo and Jennifer Davis to go ahead and present on that topic,
Representative Stephen Meeks
Unverified
56:12
if I may. Before we jump to that next topic, I have Representative Love had a
Representative Fredrick J. Love
Unverified
56:21
question. Just kind of piggybacking, because all right, so we're not actually warehousing any data, correct? Is that correct? We're just
acting as a hub okay i mean is it that is correct
Speaker 28
56:33
back back to the proposal here yes we're not
Representative Fredrick J. Love
Unverified
56:36
we're not warehousing data because i mean i i i see the data warehouse program but in actuality we're not going to be warehousing any data is that correct because i yes we're not going to make like
Representative Stephen Meeks
Unverified
56:50
a big server farm and put everybody's data all on one place everybody will keep their own data that And then you'll just have this engine that sits on top and pulls what's needed to do what analytics or whatever the case studies are.
Representative Fredrick J. Love
Unverified
57:02
And just pull everything together that's needed. Okay, so, all right, so we're, because I remember when we first began, it was like, no, we're not warehousing any
Speaker 131
57:13
data, but, okay, that's what I needed to understand. Yeah, the architecture, to pull it all into a single,
Speaker 105
57:19
a lot of folks have tried to do that. That's very, very expensive. and essentially what you would be doing is asking all these agencies to trash the investments that they've made and that's just that's not smart so essentially what we're doing is adding a layer
on top of your existing infrastructure and architecture that allows the data to talk to each
Speaker 131
57:37
other yes sir thank you any other questions before we move on follow
Senator Will Bond
Unverified
57:45
up while we've got him here and I don't know if you're about to leave But this is one thing that I haven't understood, and I ask, I think, the people from Indiana about this, maybe North Carolina or whatever, is that we've had a lot of talk about we've had issues with data sharing agreements and getting the agencies on board. I think they're coming along. Of course, my reaction has been, well, that's one thing the legislature can do.
We can tell state agencies what to do, or if we wanted to. Why don't we just cut to the chase at some point when we're ready and say, look, you will share data. You have to
do it. Here's the generalized agreement. Why is there all this negotiation and difficulty amongst the state
Speaker 105
58:27
of Arkansas over that? Well, I think Jennifer's probably better qualified to answer this question, but typically what I've found when I work with different government agencies is there's,
even though the state of Arkansas may mandate, there is a federal mandate sometimes that trumps the state mandate. So there are some federal privacy laws that are in place, in particular when you're talking about health care data. So working through those sometimes is a very difficult task. But I do applaud you for wanting to try to break the logjam because I think the benefit is immense if we could break through that in a legal, privacy-compliant manner.
Speaker 118
59:11
so but Jennifer you may be more qualified to
Speaker 137
59:17
speak to that well I just want to say I don't necessarily envy what all of you are always tasked to do or um but you know
Speaker 72
59:27
you do get more flies with honey I mean that's kind of where some of that comes from so you can tell people what to do and you know how that comes across is one thing but you know you can you know encourage the collaboration and a lot of what it is is what we've seen is you know initially
when you talk about sharing data people immediately do get an idea of a data warehouse or they think all of a sudden we're going to take all your stuff and you know especially the stuff that is highly regulated you know we're going to take all that and that's not necessarily it so I think the proof of concepts and again you have to excuse me I have a cough drop so I think the proof of concepts have shown people like what you can do when you do have that limited use case that it's not a matter of give us all of your data we want it all and we're just we're going to just keep
it just in case we want to do something with it but it has been a very concerted effort between departments to come together and say we have this need where can what can we do about it and when they see what can be done they have been more willing to want to share and come together with that and to come together to break down those legal barriers but um you know there there is the legal barriers that we have to get through there is um you know and and we're wanting to put together kind of a framework that removes those barriers so that way when the department
of education and dhs comes together and they have this need that they want to work through that we've already gotten that out of the way i mean you know the lawyers we can kind of complicate things but we've already got that part out of the way so we're really looking at removing that but it is a culture shift and a lot of it is a fear so these proof of concepts have really shown people that you know you can share data in a legal and an ethical way to get an outcome without that fear so a part of it is it's a mindset and so that's really where the things that we are continuing to
Senator Will Bond
Unverified
1:01:16
work through right now all right i still
i'm not sure that i quite i understand if we mandate it maybe it doesn't happen as quick as we want but i think patience is kind of an overrated virtue in government so uh in life really that's why you're elected and not me we should be you know it seems like if we're going to do this we are everyone should be on board i know there are legal issues to work through and all that but that some at some point maybe in the next legislative session when i'm gone there should be a mandate that it will be done by a certain timeline or my experience
in government is that just gets it just gets drug out so that i think there has to be a deadline at some point that we get these agreements done. And
Representative Stephen Meeks
Unverified
1:02:00
I'll be quiet now. Thank you. And just to follow up on that, Senator, when I first started down this path back in like 2015, one of the things I talked to the state agencies about and tried to get them to think of is the data didn't belong to them as was the previous mindset, but they were simply stewards of the data for the state of Arkansas. And I think most of them have bought into that and they see the benefit. And based upon
What I have seen, most agencies are on board with where we're going with this. If need be, the governor could just say, hey, you guys are going to do this. But thankfully, I don't think we've got to a place where we've needed to do that yet. And hopefully we never will because everybody sees the benefit to this. Although I, like you, would like to see this done
by tomorrow. But I understand there's a lot of moving pieces here. um so okay so we we don't always
Senator Jane English
Unverified
1:02:55
agree but i i do have a
tendency to agree with you we've been
talking about this you know i i went um in 10 years ago and was talking to the folks in indiana and and nine years ago i went to um maryland and they had a data sharing system i mean they do their whole budgeting process based on data that they've got. And so we've been talking about this, talking about it, talking about it, talking about it. Sometimes you just have to have the political will to say, this is what we're going to do. We're going to step forward and do it.
Because you and I, as legislators, we can suggest and all, but it really takes, it almost takes administration because we're not in charge of agencies. That's the administration side. that's so sometimes it's you just got to figure out we're going to do it and figure out how we do it any last questions or comments before we proceed
Speaker 144
1:04:04
good afternoon i am robert magoo i also
Speaker 145
1:04:07
serve as deputy chief dead officer along with auditor In my other role, I lead the team that's building data warehouse and analytic solutions for Arkansas State agencies since 1997. I'm Deputy CDO along with Audita and the data warehouse lead for the Department of Transformation and Shared Services Division of Information Systems.
One of the responsibilities of the Data and Transparency Panel, or DTP, from Act 912 of 2017 was the development of what was called a feasibility and cost study on a statewide data warehouse program, which we sometimes just call the data warehouse study for brevity. Based on my role and my experience with DIS building a lot of large-scale data warehouses, some of these are statewide for one program, like multiple agencies for one area. Some are for an entire agency of multiple areas, but none know the scope of all agencies for all
areas. I took the role of lead researcher on this and coordinated with the different DTP members for input and feedback and kind of pulled together the final report, which you have as Appendix C in your packet. This goes into a lot more detail than I'm going to talk today, and even that's summarized, but I'm happy in the future to delve in more detail on anything we need to. this was about a two-year process from when the CDO office was established in 2017
there's a lot of research done the final draft was presented to the data and transparency panel in the April meeting there's a feedback and review period of three months that was all incorporated and the DTP voted to finalize the report in the July meeting which is this final published version that you have. The general flow for the study started with the needs. If we're going to have a solution, we need to know what the needs are.
These were laid out in a lot of different acts, starting with 1282 of 2015, 912 of 2017, and even a couple in the recent legislative session, such as 936 and 943. The executive agencies also had a lot of input when we did our CDO listening tour and when we did the GAP analysis report, which is also a published deliverable that's available where we kind of assess the data needs of all the different executive agencies. So we took all those different needs, kind of developed a taxonomy,
pulled them together, and organized them into just six key areas. And those are listed here as accessibility, which is transparency, which could be to the public, and self-service data access and delivery. That's what's really provided a lot of value in some of the jurisdictions and other states we studied is that self-service access. It brings operational efficiencies and helps policymakers and decision makers. Sharing while enforcing privacy, security, and compliance. So that's sharing the right data with the right audience for their role.
Enhancing data analysis capabilities still allows to support better data-driven decision-making. integration to enable cross-agency analysis and to reduce duplication data and effort and that kind of sort of goes back to the Rosetta Stone and how we can start linking these data sets between agencies and breaking down silos timeliness in order to enable real time assessment support and alignment of programs and departments and data quality so that we can
manage data as a state asset and that's an important foundation for all of this particularly as you get to artificial intelligence machine learning, things like that, the data quality is critical. After the needs were identified and sort of summarized into these six key program objectives, we did a lot of research. We talked to 15 different U.S. states. There's an Australian State, New South Wales, Australia, that has a very impressive program, and a large Canadian
federal agency. The Treasury Board Secretary of Canada is doing a really good job with that self-service data access and delivery. This was to interviews with their CDOs, CIOs, or their data personnel, and review of their published research. We also bounced this against published research and analyst interactions with industry analyst firms such as Gartner and published best practices from data professional organizations
like the Data Warehouse Institute and the Data Management Association. So after all this research and we studied these approaches, we applied the successful examples from all these different states to our identified needs. And this was kind of a mix and match of what worked really well. We saw Indiana was a really good example, pretty holistically, of a lot of these things. Massachusetts did a really good job of data sharing compliance.
They all had their kind of bright spots where they were excelling, so we tried to take sort of a best-of-breed approach so that we can kind of start in 2019 with the current best practices. The feasibility of each of these was assessed based on assessment of success or smaller scale implementations here in Arkansas, prototyping and proofs of concepts such as the ones we've discussed today and previously. And finally, the cost and benefits for the program components were estimated and next steps towards implementation were recommended.
The study in great detail identifies 11 key capabilities and approaches based on all of that research, recommended for addressing the program objectives. These are covered in the slide with the different layers. This is a layered approach between data providers and data consumers to show how data can move in a governed manner between the data providers and consumers. The study was called the Data Warehouse Study, and I've been building data warehouses
for a long time. That is what we used to do in the 90s and the 2000s where you take data and you make a copy, and you put it in a centralized repository. And it's effective, but it's not always the optimal approach, and there have been a lot of evolutions. So we've talked a lot more recently about the DataHub approach, and that's sort of an evolution of what the Data Warehouse really is today. And the Data Warehouse is really purely analytical in nature for looking at historical data and supporting decisions. DataHub brings in some additional technologies and modern
best practices to also help with operational needs and that gets us to continuous program alignment and these things that can help us not just see what happened but to kind of shape optimal outcomes going forward with artificial intelligence machine learning and things like that not covered in detail in this report but critical is data governance that's a whole other subject to itself it is covered a bit in the gap analysis and that's a continuing topic of discussion with the Data and Transparency Panel and
Speaker 153
1:11:38
will really be fleshed out with the agency data officers.
Speaker 145
1:11:49
Along with the approaches from the other state, we looked at the benefits from the other state. We already heard from Indiana last time about their $40 million return on investment over an 18-month period. The Arkansas recidivism POC mentioned earlier today has an estimated return an investment of $8.1 million in a three-year period. There were lots of other examples, including Texas unemployment insurance fraud reduction of $71 million over a three-year period, and that was similar to what we were talking about earlier from integrating the
corrections data with the unemployment insurance data to reduce unemployment insurance fraud. And a lot of other non-financial examples, Indiana, the very first use case was around infant mortality, and they had some pretty successful results reducing infant mortality. They worked on highway crash safety. A lot of states are using integrated data and analytics to address the opioid epidemic and to improve education and workforce development policy and decision support. Okay.
Representative Stephen Meeks
Unverified
1:12:56
Any ideas you all deserve? I just wanted
Speaker 72
1:13:00
to go ahead, and for those
of us in the room that are not as smart as Robert I just wanted to kind of you know pull this back together to say that you know a lot of the work that the CDO office has done you guys had asked us I think in the beginning what work had been done and they've been working on this report and a lot of this not the report specifically but this stuff for for a very long time and analyzing all all of that data so it's a lot of data that came into that and really this is just a way that you know we can look at those current initiatives and kind of use that rosetta stone approach to kind of
you know help agencies and departments really have their data flow because you know all all the departments you know in the state are really making the best decisions that they can right now but it's a matter of really being able to put those departments together to make better decisions you know one of those examples and and you know another use case that we hope at some point in the future um you know we'd love to come and talk to you guys about it is we're currently working on an education and workforce proof of concept and how that started is you know um at that time it
you know the transformation had gone in but you know k-12 um higher ed workforce did we have pre-k okay as i said i wasn't sure but you know we had you know um career ed these entities came together and talked about you know what issues that they're having because you know every department has an overlap into another department and so they came together and I think it's more specifically higher ed and workforce but they're looking at like a skills gap analysis and next week Robert has set up some meetings with the stakeholders to review the findings of that
and again like I say I hope at some point you know we'd love to come back and show you that but it's another example of what happens when those agencies and departments can come together to work together, to share data, to come up with a common goal that they need to reach. Because, you know, that skills gap analysis really helps everyone. And
Representative Stephen Meeks
Unverified
1:15:07
so that's really how that kind of pulls together. And on
the study, did you come up with a reasonable implementation timeline,
how long it would reasonably take to get all these, you know, start to get some
Speaker 144
1:15:24
of these online for our state we didn't look at the timeline without knowing the
Speaker 145
1:15:30
details of the scope it depends on a lot of factors the order is important and some of these things should be started before others particularly master data management that rosetta stone because until we can link the identities it limits what else we can do other capabilities can tear in
over time and it really depends on the business cases and the needs how much can you know this is infrastructure capacity it's compute without knowing the exact data volumes and the exact compute needs it's really hard to work out some of these finer details okay what do you
Representative Stephen Meeks
Unverified
1:16:03
all see as the next steps in that process though how do we start how do we make sure we keep going forward from
Speaker 72
1:16:11
here um we are continuing to um you know work with the stakeholders work at the data transparency panel members to identify those issues and work through them much like the education to
workforce and i go i do want to go ahead and address your question i know that you've i think started to answer to ask it a couple times about the funding and i will say that the funding that is listed um in the final cost feasibility study now keep in mind that was a pre-transformation so we are currently um you know reviewing the cost that would be associated with something like this with our executive leadership because, you know, the transformation has offered a lot of benefit already in the short time to break down some of those barriers. A lot of agencies that
looked at interagency, you know, agreements that made it more difficult are now in one department, which kind of, you know, just helps naturally in that idea of data sharing. It brings additional people together and things of that nature. So we are reviewing that number with our executive leadership to determine what really is a post transformation accurate cost for that amount i don't necessarily have that number today because it is under review but i didn't want your question to continue to hang out there and think that we weren't trying to answer it so well and
Senator Will Bond
Unverified
1:17:28
i may have just gotten confused but the the which is not your fault that
but in this report doesn't say that's where if you it's got dollars in here somewhere right it I thought I highlighted
Speaker 87
1:17:42
it on the handout or the actual cost feasibility
Senator Will Bond
Unverified
1:17:46
report so if you look at page 29 right so that the annual
and I may be I was reading too quick the annual cost for
establishing supporting state data-wide or estimated 3.9 million a year for years one through five and then you go down to total annual costs for other statewide data warehouse programs studied range from 8 million to 9 million per year so I think kind of what I was wondering is that is it the 4 million plus 8 or 9 million a year I guess in something previous we were given I think it It was $8
Speaker 145
1:18:31
million. The $8 million and $9 million, those are from other states. Those are from Indiana, Michigan.
Those are just examples of what these types of programs are running in other states. So the total is $4 million per year? Yeah, the total from the study was the $3.9 million per
Speaker 72
1:18:49
year. Right, and again, keep in mind that that is a pre-transformation total that we are currently reviewing. So I can't say that that number is still accurate. but at the same time, I am not the financial wizards of my department, but keep in mind that
we are reviewing that to determine in light of the transformation, what opportunities have arisen such that it would affect this because this was done prior to the transformation. So, right. I
Senator Will Bond
Unverified
1:19:20
understand the hedge. I'm not giving you a hard time about it, but you're saying that you need an appropriation for four million dollars a year reoccurring to put together what we showed earlier the the hub is that right to put that together
Speaker 170
1:19:37
with the hub in the middle or no again like
Speaker 72
1:19:42
i say based on the transfer that
number might have gone up or down i don't i can't say accurately that the 3.9 million is the amount so i understand no i
Senator Will Bond
Unverified
1:19:52
understand so i just don't want to lead you into saying one way or the other the reason i keep going back to the data sharing agreements seems like it would not be smart for the legislature to fund this and not know the data sharing agreements were already entered into or timely would be entered into it would just seem like bad business and so i differ a little bit in that i think we can tell agencies in their budgets you will have a data sharing
agreement by such and such a date or you don't receive such and such so but I'm a little more aggressive and you know pointing at
the agencies maybe than some who have more influence in the governor's office and the challenge to doing something like that would be are we going to who
Representative Stephen Meeks
Unverified
1:20:33
are we going to tell them to share with is it just going to be a blanket statement you will have a debt of sharing or are we
going to say you're going to share with a b and c and how do So, you know, I guess the devil could, it's like a lot of stuff we do. It's simple on concept, but the devil's in the details.
Senator Will Bond
Unverified
1:20:51
No, I agree, and I'm not being adverse. I'm just saying I think I agree that it's a very important thing. I'm just saying if we're going to fund it, whether it's $4 million, $8 million, $12
Senator Jane English
Unverified
1:21:04
million, then we want to know that it's going to happen. It's going to happen. I agree. Yeah. So if you're batted around, and I think people
have come up with recurring cost agreements or whatever, and we've had those things before where the agency has to pay for versus funding the whole thing.
What is that $3.9 million? What does that in the $2.8 a year? Is that just you all doing it without agencies have to be paying every time they turn some data in or ask
for one of these cases? you know
Speaker 72
1:21:44
because the department of or now division of information systems excuse me is a cost recovery agency that does present some of those challenges and again you know both Robert
and I are not really necessarily in that position to address the financial area like that because neither one of us serve as the CFO of our division or our department now Ann Purvis is here if you want to ask her. She does serve at that at our cabinet level, but I don't feel that I'm in an appropriate position to address too many financial or chargeback type. I just kind of wanted to know what does this
Senator Jane English
Unverified
1:22:18
$3 million, so Anne, if you wanted to come up, that would be wonderful.
Representative Stephen Meeks
Unverified
1:22:27
And once you get settled in, if you would
Senator Jane English
Unverified
1:22:30
just introduce yourself to the committee. So I'm kind of looking at what does that $3.9 million
include, and what is the $2.8 million?
Ann Purvis
Unverified
1:22:40
What is that? My name is Ann Purvis. For this purpose, I'm Chief of Staff for the Department of Transformation and Insured Services. It is a blend between FTE time within the department as well as probably the system approach to the integration and to help with us to create the data hub itself.
But because a division of information system is a cost recovery department, it becomes a challenge to make an obligation of this magnitude without knowing that the departments would actually utilize it so you would not under-recover a cost. In other words, you would basically end up paying out more money than you're taking in. That creates a problem for the department. So one of the things we wanted to do is step back and really look at this from maybe a different angle, because one of the things I've been doing for the last 17 years is doing federal cost allocation modeling
across different departments, including within DF&A. So that's the challenge that we have, is not to over-obligate the department in such a way that it cannot recover that cost. And most of the money, if you look at the, comes from federal programs. and we said this is the only way for us to recapture that federal dollar back into the department for these type of activities and so we just need to be very mindful of
Speaker 179
1:23:57
that as we design this moving forward
Senator Jane English
Unverified
1:24:00
so i guess that my concern would only be is i i hear some hesitation from agencies because they don't want to have to
do a cost recovery thing but as you're saying but
you remember some years back where um there was the cost recovery that dhs was paying to dis and it was for the wrong funds and we ended up having to pay back some 30 million dollars
and you remember that i am very familiar with that
Ann Purvis
Unverified
1:24:26
i figured you probably would be so that's one reason i'm very sensitive to being able to put not to put the
Senator Jane English
Unverified
1:24:33
state in that position again right and so i don't that's why i'm kind of trying to figure out
is this 3.9 million dollars is that so we could do it or and then the 2.8 million is that a recurring cost that without the cost
Speaker 181
1:24:47
recovery or is that in addition to cost recovery that is that is currently what the projected cost would be
Ann Purvis
Unverified
1:24:56
overall how we actually recapture that cost back becomes the question and whether or not we would have enough use cases in order to capture that cost becomes another question so that's and not whether the agencies can afford it maybe the third question is because this would be an additional cost currently not within their budget structure but if they're going outside their department for this type of analysis or this type of of integration of data, then it makes sense.
So that's one of the things, you know, Jennifer is absolutely right, Ms. Davis, in that the transformation has created this unique communication network across the departments where we're able to talk with 15 different CIOs, talk with 15 chief of staffs and say, okay, this is what we're looking at. Let's go back down into your divisions and find out where are our best use cases for this. And a lot of us, you know, because I come from the Department of Health and several different other places in government, We already had data use agreements. We were already sharing birth and death records with Medicaid
to make sure they could clean up their records on an ongoing basis. So this isn't new to all state agencies, but they're absolutely right. Some of them kept that very protected, and so we need to be able to share that information across all
Senator Jane English
Unverified
1:26:07
those different departments now. Well, I think for us as legislators, because I'm all for this, but I think being able to, and as we
go hopefully shortly, to define what that cost is because the fiscal session is coming up pretty soon. It would be nice to have some idea
of whether we need it to be, as Senator Bond said, whether we need to find a way to appropriate those dollars so that we're not funding things. But if we're going to do it, we need to do it. But we need to figure out what that cost is. And I just would like to know what that 3.9 is. Is that bringing back costs? But I just, in my mind, I remember this whole thing with the cost recovery about all those dollars, those federal dollars that we ended up having to pay back, and we don't want to get in that situation again.
And I know that cost recovery is a problem for a lot of
Representative Stephen Meeks
Unverified
1:27:01
agencies. Absolutely. And I think ultimately, obviously, I'm only
one of 135, but in my mind, the best approach is what we need to do at DIS because you are a cost recovery agency. and this is not designed under that model to work as a cost recovery, that ultimately what we're going to need to do is create a separate data division that's funded from general revenue.
And then as the agencies see savings because of this effort, then we can make those deductions out of their budgets, out of the appropriations of the agencies as that occurs. but I think ultimately they need to have a stable funding source and not be looking for the agencies to try to try, because that just complicates things in my mind. And Representative Minx, the only
Ann Purvis
Unverified
1:27:53
thing that I would offer is for us to continue these conversations, because I absolutely want to grab that federal dollar,
and the only way to grab that federal dollar is through some sort of allocation methodology that the federal government will agree to. So, you know, if we didn't do this approach, then you'd have the bigger agencies like DHS that may use it a lot. We can't capture that money back to support the operation. So it sounds like we may
Representative Stephen Meeks
Unverified
1:28:18
need to look at some sort of hybrid model then perhaps. I
Ann Purvis
Unverified
1:28:21
think it's worth a quick look. Okay. And
Representative Stephen Meeks
Unverified
1:28:24
that's what we're counting on, y'all, to make those, you know, come up with those proposals and bring them to us so we can make the best decisions possible.
And like Senator English said, I'd like to try to have something in place so that we can move forward. You know, for those of our colleagues that are still sitting here, I think one of the most fundamental things we need to look for during the fiscal session is the presentations we've heard today have been from our deputy chief data officers because we don't have a data officer. That position is currently empty. And, again, I think that needs to be a general revenue-funded position.
And so I think one of the things, and y'all may or may not agree, but I think one of the steps that we need to look at doing is we need to find that person who's going to be our lead on this, whether it's in-house, you know, whoever that chief data officer is. Because that's going to be a big question mark on moving forward is who that person is,
what their vision is, and so on. Does anybody have any other questions for the panel, or does our representative Brown?
Speaker 56
1:29:33
Thank you, Mr. Chair. Ms. Parvis, I believe you answered this question when you were responding to Senator English, but maybe you could elaborate just a little more. I'm curious if you have brought all the secretaries together or if you've had an opportunity to present this to all the secretaries at one time and encourage their collaboration and get them excited about saving money for their departments
since this restructuring is supposed to save the state money. So we need them to buy into this and get excited about it. Have they all been together in one spot? The secretaries have met on several occasions, as well
Ann Purvis
Unverified
1:30:19
as the chief of staffs, communications officers, the CFOs for each one of these 15 departments. and also we're gathering the CIOs now. We wanted one point of contact for each one of the departments so we can talk about some of these issues.
But it's both, you know, one of the things that I look at it, I'm more of a business owner approach to this and have the value with these type of efforts. And so, you know, it's not only a technical solution, but it's also a business solution of government. And so bringing those two things together, and that's one of the priorities that we're having at our next chief of staff's meeting, the operational end of some of the departments. As I say, these are the efforts that are going on right now. I give a briefing on the Department of Transformation and Shared Services on each one of the different areas
and how we're trying to become more efficient in serving those different other state departments. This is on the topic of the next meeting. Thank
Representative Stephen Meeks
Unverified
1:31:21
you. All right. Any other questions, comments? Okay. Do you all have anything for us before we? All right. Well, again, we definitely appreciate your efforts on behalf of the state, and I hope I think I speak for all of us. This is definitely something we want to see happen sooner rather than later, and you have our confidence that we'll get there.
But I think, colleagues, we're looking at
October 9th for our next meeting. I think one of the things that we want to
Senator Jane English
Unverified
1:31:53
do, we can't probably do it as an official meeting of this
task force, but Axiom has offered to give us a big picture look look at what the world looks like out there with data and all so we could have
and anybody could come task force person or public or whoever wanted to come transparency panel so we'll set that date up and we'll get Blake to kind of make an announcement and then whoever wants to attend would be more than happy to have you be there and attend probably won't be an official task force trip but I think it's important for most of us have absolutely no idea what this big world of data looks like and and you know we're kind of operating in this never never land so I think
from my standpoint I know that I would like to see something like that and see what what we're actually talking about how it is how it is being used so we'll wait for the the date sometime and I'm hoping probably by the middle of October we'll be able to do that. All right. I
Representative Stephen Meeks
Unverified
1:33:06
appreciate everybody being here. I think we've got a bright future as a state. And with that, we are adjourned.
Agenda
A. Call to Order
B. Comments by the Co-Chairs: Senator English and Representative Meeks
C. Consideration to Approve August 14, 2019 Meeting Minutes
D. Discussion of Data Catalog efforts and Data Cleanup and Enrichment. - Ms. Yessica Jones, Director, Department of Information Systems
E. Feasibility and Cost Study Report - Ms. Yessica Jones, Director, Department of Information Systems
F. Other Business
G. Adjournment
Documents
| Title | Type | Pages | Source |
|---|---|---|---|
| Agenda — DATA SHARING & DATA-DRIVEN DECISION-MAKING TASK FORCE, Sep 11, 2019 | Agenda | 1 | Official source ↗ |
| Exhibit C Draft Minutes | Exhibit | 2 needs OCR | Official source ↗ |
| Exhibit E Feasibility and Cost Study | Exhibit | 30 needs OCR | Official source ↗ |
| Handout D State Data Catelog | Exhibit | 16 needs OCR | Official source ↗ |
Speakers
Representative Stephen Meeks
Unverified
Senator Jane English
Unverified
Speaker 14
Aditha Karkara
Unverified
Speaker 25
Representative Fredrick J. Love
Unverified
Speaker 28
Speaker 56
Speaker 58
Speaker 62
Speaker 6
Speaker 72
Speaker 55
Senator Will Bond
Unverified
Speaker 85
Speaker 77
Speaker 92
Speaker 95
Speaker 41
Speaker 105
Speaker 115
Speaker 119
Speaker 131
Speaker 118
Speaker 137
Speaker 144
Speaker 145
Speaker 153
Speaker 87
Speaker 170
Ann Purvis
Unverified
Speaker 179
Speaker 181