House Defense & Veterans Affairs Committee, February 3, 2026 Video: https://vimeo.com/1161540502 (a timestamp t plays at https://vimeo.com/1161540502#t=s) Page: https://georgiacommons.org/meetings/2026-02-03/house-defense-and-veterans-affairs This is a machine transcript of the General Assembly's video, made by a speech-recognition model. It is not an official record and can mishear words, names, and numbers. Speakers are not identified. The video is the record: every timestamp links to the moment it plays. [0:01:42] Hey, if we could clear the room, uh, for the next meeting, we would appreciate it. [0:03:12] All right. All right, good afternoon, everybody. This is the House Defense and Veteran Affairs Committee hearing. If you're here for something else, you're welcome to hang around. Uh, we're gonna go ahead and get started. It is just a few minutes after 2:00. Uh, at this time, I'm gonna call on Representative Ballard to give our invocation. [0:03:32] If you'll bow with me, please. Heavenly Father, thank you so much for allowing us to be here together today to do the work of the people of Georgia. Please help us keep in mind, um, what we are supposed to be doing, and help us to focus on our citizens and our veterans who have given so much for us. Um, in Thy name we pray. Amen. [0:03:57] Thank you very much. All right. Um, this afternoon we are not gonna be considering legislation, but we're gonna have an opportunity to, uh, hear about an important topic. Uh, before I, I do get started though, I just wanna make a couple of, uh, quick recognitions. Uh, we've got Commissioner Patricia Ross, our Commissioner of Department of Veteran Services here with us. We appreciate that. And we've got our Deputy Adjutant General, Joe Ferraro, in the back as well. We appreciate that. Uh, and now to introduce our topic, I would like to bring, uh, Major General Retired Grabowski to the podium to, uh, introduce the folks that are gonna be giving our presentation today. [0:04:32] Uh, Major General Grabowski was a, is a former commander of our Air National Guard, and currently serves as the Georgia State Director for Selective Service. So we appreciate all the service you've given to our state and the nation, sir, and we would appreciate, uh, a quick introduction of our guest today. Well, Chairman, [0:04:53] Words to this distinguished body today. And, uh, thank you for putting me on the spot as well. I thought I was gonna sit here quietly with Commissioner Ross. Uh, but as a 37-plus year member of your Georgia National Guard, when these gentlemen came to me and said, "Hey, we've got something that we'd like to talk to you about," uh, as a Air National Guard commander, uh, I was intrigued. Um, by way of introduction, I'm gonna let them introduce themselves, but I've known the gentleman to the left for almost two decades. [0:05:23] We served in combat, flew in combat together, and, uh, we sincerely appreciate, uh, your time today and allowing them to, uh, brief this program. And with that, uh, Bigfoot, I'll turn it over to you. [0:05:37] Thank you very much. Um, we did have slides. Are there no... Did w- no slides come up? [0:05:43] Well, um, so every, every- [0:05:45] Oh, perfect. Uh, great [0:05:47] ... but we can also put them up on the screen. [0:05:48] Okay, great. [0:05:49] But everybody does have the handout- [0:05:50] Yes, sir [0:05:51] ... to start, um, referencing that. [0:05:53] You bet. Um, well, most importantly, thank you. Uh, thank you, Chairman. Um, thank you, Representative Ballard, for, um- [0:06:01] Pull that mic just a little bit. [0:06:02] You got it. How's, how's that? [0:06:04] There you go. [0:06:05] Okay. Um, thank you everyone for the opportunity to, um, to talk about a topic that's really important to, um, to both myself, um, and to Marcus, um, who will introduce himself in a second. It's an honor. Um, my name is Nelson Rula. Um, I'm a prior enlisted airman, um, and previously retired Air Force colonel. I've had the privilege in serving in the 116th Air Control Wing, um, and then later got to command, uh, a JSTARS squadron here in Georgia. But today I'm here for a different role. Um, today I'm, I'm, I'm focused on helping service members prepare stronger VA disability claims paperwork in a way that is dignified, accurate, and most importantly, absolutely free to the service member. [0:06:48] Alongside myself and my colleague, Marcus Smith, together we co-lead VetScore, and our mission is very simple: ensuring medical records are accurate before a claim ever reaches a VSO or the VA so the system can work exactly as intended. We believe the VA disability system works best when it receives accurate documentation from the start, and that's the very narrow gap that I wanted to talk to you about today. See if I can, uh, I should switch slides. [0:07:22] Today, we believe many service members navigating the VA claims process face four imperfect paths, right? They can go through a VSO, very honorable American Legions, VFWs, uh, or the like, but often we see that the, the, the demand outsees the capacity They can pay private attorneys, accredited agents, medical providers. But those costs become a barrier for many families. Some try to go it alone, will go online, Reddit, Facebook, TikTok. [0:07:58] Uh, that is good advice, but it comes at a very risky price down the road. And so what I wanna focus on is what we believe is the underlying issue. [0:08:13] Slide, please. There we go. If we step back and ask what's really driving the many claim challenges, we find two root causes routinely occur. First, certain treatment records, or in our parlance, medical records, they're often missing, they're incomplete or inconsistent. And second, the claims process itself is complex. We know it's rule-bound, it's evidence-based, but sometimes it's difficult to navigate without help. [0:08:47] When those two realities meet, the outcome is predictable: delays, rework, and sometimes avoidable denials. We believe this creates stress for the service member, their families, and for the amazing men and women that work at the VA. [0:09:12] When documents gaps occur early, the downstream effects are predictable. We see more exams, we see more appeals, and we see longer wait times. And that's, that workload goes across the system, and this is what's frustrating for service members and sometimes the VA. We want to be clear, this isn't about fault anywhere. It's about timing and information quality. When documentation improves earlier, the entire system works more efficiently. [0:09:46] This is exactly what I'd like to discuss for the next 15 minutes. [0:09:55] Another reality we observed is the financial barrier associated with the claims preparation process. Some families choose to pay for help navigating that process, and that can be a very rational choice for many, but it can also create hesitation or delays for too many. Families often, often carry that burden quietly, and in some cases, that cost prevents service members and veterans from seeking help or seeking counsel. [0:10:25] Our model removes that barrier entirely. The service member never pays. No hidden cost, no fees, not today, not ever. Access is sponsor-funded, so the focus stays on preparation and documentation quality rather than contingency fees or any sort of upselling. We don't think documentation quality should ever depend on a family's ability to pay. And by removing any financial barrier, we help ensure the emphasis stays where it belongs, on getting the record right the very first time. [0:11:03] We work with service members long before separation or retirement, long before separation or retirement, to help them assemble clear, complete, and provider-ready documentation. And we do this with a combination of human expertise and structured AI-assisted checklists that help translate lived experiences into documentation that providers in the VA system can easily understand. Importantly, we work in sync with VSOs and the VA. [0:11:39] We are not trying to replace anybody. We are helping service members arrive better prepared for their first meeting with their VSO or the VA. [0:11:50] The result is much more accurate documentation on the first pass. Across about 700 claims to date, including conditions like migraine, tinnitus, sinusitis, PTSD, depression, we, we are exceeding 92% first pass rate. And here it's important for us to be clear by what we mean by accurate. We don't mean more claims. We don't mean higher ratings. In fact, sometimes it's the exact opposite. We see fewer claims and even lower ratings at the end. [0:12:25] For example, a service member might talk about having migraines. When our humans and our software find a disconnect or misalignment in that, they'll go back to the service provider, and the service provider may properly diagnose them with tension headaches, which is very different than a migraine. We consider this a win, even if it results in fewer claims or a different rating or a lower rating. Our goal is to align what is claimed with what is experienced and medically supported so that the record reflects a service member's and a veteran's true condition. [0:13:09] We believe this challenge exists at real scale. Each year, roughly 200,000 service members transition out of service. On average, veterans receiving compensation are managing documents for multiple conditions, often settle-- several per person. For those who do receive benefits, those payments provide meaningful support as families adjust to civilian life. The takeaway here is not the dollar figure at all. The takeaway is that a small improvement in documentation accuracy can have significant impact when applied across thousands of claims throughout the year. [0:13:49] If it's okay, I just wanna take a small moment to introduce our team. We built VetScore with people who understand both military service and complex systems. Together, we've served, we've fought, we've commanded, we've built multimillion-dollar companies with several hundred employees. We've worked on front-end, back-end software from startups to Fortune 500 companies. The common thread in this small team is that it cares about getting this right and operating responsibly in a very sensitive space. [0:14:20] We're not large. That's intentional. It helps us keep us focused and accountable. Two of us are veterans, and other two teammates were raised in military families. So we built this with firsthand understanding of how these records affect families over time. [0:14:41] The next two slides are gonna talk about the actual software. Our approach is structured and methodical. We guide service members through a step-by-step documentation process while they'll s- while they are still serving, long before they separate or retire from service. This includes helping them organize their records, identifying gaps, and preparing documents that medical providers can review and validate. We use structured checklists and AI-assisted prompts to keep things consistent, but humans always remain throughout the entire process. [0:15:18] We don't make any medical judgments. We don't influence any decisions. Our role is simply to help members present clear and complete information so the system can do its job. A simple analogy is TurboTax, if you're familiar. TurboTax doesn't decide your taxes, and it doesn't replace the IRS. It helps you prepare information clearly before your submission. VetScore plays a similar role in preparation. [0:15:50] We help service members organize their documents, and then it goes to the VSOs or to VA for formal review and submission. Our job is to improve the inputs so the system can do its job as intended. [0:16:07] The VetScore platform that you see is the backbone of our document support. It organizes a service member's records into structured, reviewable claim issues so gaps are visible early. Each condition is reviewed for medical evidence, procer- d- proper documentation, and internal consistency. The platform is designed to mirror the VA decision logic, not to replace it. We do not adjudicate, diagnose, or make promises or judgments. We simply help improve the quality and completeness of what is submitted. [0:16:41] By catching gaps early, we reduce denials, rework, and supplemental claims. The result is more accurate documentation and fewer preventable errors along the way. We believe this allows the VA reviewers to work with the stronger inputs from the start. [0:17:02] Now, I, I wanted to share this slide to show that this work is already operational and attracting thoughtful partners. To date, we've supported documentation across about 700 claim issues, and as I mentioned, are seeing about a 92% first pass rate. For us, that's a real quality metric. It tells us whether service members are arriving better and well-prepared. We built this gradually and responsibly, not trying to scale too fast. We've been lucky enough to receive some philanthropic and institutional partners that have chosen to support early pilots and governance. [0:17:39] On the compliance side, we were very lucky to have Holland & Knight to advise our HIPAA-aligned privacy and governance structure so we can operate smoothly in a sensitive area, and they're doing this all pro bono. As part of this broader work, I was appointed, uh, I was appointed as Stanford Veteran Fellow, which gives us significant research and policy lens as we try to scale. Stanford has also offered the opportunity to stress test our model through learning loops and research assistance. [0:18:12] So at this point, we think the question isn't whether or not the VetScore approach works. We think the question is how to scale it responsibly to more service members and eventually to veterans and to the system so everyone can benefit. [0:18:34] We've really tried to keep our model simple. The end user is always the service member. Our intent is that the service member never pays a dime. Today, we've primarily been self-funded with early support from a small number of partners as we've built and proven the model. Going forward, the sponsor model is the path. Philanthropic foundations, grants, military units, and law schools with veteran centers. [0:19:06] That's the structure we're here to hopefully start a conversation with you. A disciplined, measured way for Georgia to sponsor access for Georgia service members. We're early, but we're operational. We've been very careful about scaling responsibly. [0:19:29] This slide will outline how we will-- how we aspire to roll this out. You'll see the approach is phased and measured. We built the initial MVP and prototype ourselves and used that phase to learn what works before ever seeking help. We started small by learning and expanding only when the outcomes justified it. The first phase focused on refining the workflow and ensuring documentation quality and data protection standards were strong. The second phase uses defined pilots with clear metrics, documentation quality, provider usability, and participation satisfaction. [0:20:08] Only after these metrics are met do we scale to anything larger than what we've been doing. We think this structure allows a state partner to support a defined number of service members, to see results transparently, and to make evidence-based decisions about continuing or not. It's designed to be accountable from the very beginning. This isn't about growth at all. It's about responsible growth. [0:20:39] Here is the current landscape as we see it. There are strong actors in this space. There's no doubt, and we generally value the role that everyone plays. At one end, you have law firms and medical consulting companies. They offer high expertise, but can be expensive and are typically engaged later in the process, much closer to separation or retirement. You also have accredited agents and VSOs and many other amazing organizations. They are absolutely essential to the system and do important work filing claims and advocating for veterans. [0:21:14] But they're often operating under heavy caseloads and time constraints, again, very close to when a service member is thinking about separating or retiring. Most people don't walk into their first VSO meeting until eighteen months, twelve months, and some sadly, six months or less before they separate from service. And our contention is that is not enough time to have the records clear and accurate. We talked about the social media and the self-help, Facebook, Reddit, all of which are accessible but can be inconsistent and sometimes just have one wrong, one wrong answer that will lead down to many months of frustration. [0:21:56] The gap Vets- VetScore focuses on is not replacing any of these important players. It's timing and translation. VetScore's role is very narrow and very upstream. We focus on documentation preparation long before submission, so service members can engage and walk into their VSO or VA- VA office with the most accurate records. [0:22:25] We are trying to help service members arrive better prepared. [0:22:32] This slide is about transparency on cost and sustainability. Today, supporting a service member requires staff time to review records, organize documentation, and prepare the material for the review. We track that time carefully and work to reduce it through better tools and process refinement, and we are getting better every day. As the workflow improves, the cost per member goes down while the quality increases. The sponsor model now allows a state or an organization to support a defined number of service members at a very predictable cost. [0:23:05] That makes budgeting straightforward and outcomes measurable. The goal is sustainability and ef- and efficiency, not growth for growth's sake. We're trying to be careful stewards of every single dollar. [0:23:24] This slide is similar, and it just shows what a sponsorship delivers. A sponsored Georgia service member gets structured documentation support while still in uniform, while their records can still be easily remedied. They receive guided preparation, so their documentation is organized and provider-ready when they choose to file. Once again, there is no cost to the service member. From the state's perspective, sponsorship funds a defined number and a level of support. [0:23:54] In short, sponsorship turns early preparation for the member and cleaner documentation entering the system. [0:24:04] Just one more slide on the phases, because we think it's really important that you know that the work is around clear milestones. Each phase has clear checkpoints, so progress is measurable to everyone. The build phase focuses on refining the workflow and quality standards. The proof phase uses defined pilots to confirm documentation quality and usability. We use learning loops at each stage, measuring results, adjusting, and improving before ever moving forward. Only after those checkpoints are met do we move forward. [0:24:36] And conversion simply means transitioning successful pilots into sustained programs. And each sta- at each stage, there are significant decision points. If outcomes don't justify continuation, the model allows for adjustments before expansion, and the structure designed to protect both the service member and the public resources. The key point is this. This is measured, data-driven progression with clear checkpoints at each and every stage. [0:25:05] Allow me to make this personal for just one minute, or maybe three. This work starts at home for us. All of our founders' fathers served in Vietnam, and their service shaped our families and our lives. This photo was taken in, in Hue, Vietnam about three or four years ago. We were fortunate enough to take a trip with our fathers and return them-- and return to Vietnam. We walked the ground they once walked. We looked over the airspace they once flew and the waters they sailed, and listened as their sh- as they shared stories about their time. [0:25:39] It was a chance to see their service through their eyes and through their lens, and to better understand the parts of their lives that were shaped long before we ever knew them. And standing there with them, we realized something very important. We saw memory, meaning, and the quiet, enduring cost of service. I think many of us had our own questions answered about why our fathers behaved the way they did. [0:26:05] We grew up witnessing how service can change a life, shape a life over decades, often carried quietly without complaints. In my own family, I saw how hard it was to translate lived experiences into the kind of documentation the system could actually recognize. My father, like many Vietnam veterans of his generation, didn't go to sick call. Years later, when paperwork and proof mattered, the records told an incomplete story. The reality of his service and its impact simply wasn't captured anywhere. [0:26:39] In the end, he spent over $30,000 out of his pocket to have experts review records and assemble the documentation needed to reflect how he actually lived. [0:26:51] $30,000. Most veterans can't afford that, and they shouldn't have to. That experience is a key reason why VetScore exists. [0:27:03] We built this deliberately. We bootstrapped ourselves, our early development, and scaled step by step, earning validation along the way, from William & Mary's, from William & Mary to PenFed Foundation's Veteran Acceleration Program, and now through Stanford University. Now we're asking Georgia to help us take the next responsible step, smartly scaling with evidence. [0:27:29] And Georgia's personal to me, too. I was lucky enough to command and fly with the JSTARS community here alongside airmen who called this state home. I sat with Georgia families during deployments and homecomings, and when you serve alongside people, their home becomes personal to you, too. I don't have to tell anyone in this room, and nor would I dare, that Georgia carries a strong military tradition and a real commitment to those who serve. I think best in the nation. [0:28:01] We also think this makes it a natural place to test this model carefully and measurably, and if it works, to help set national standards for how we support service members before we leave uniform. That's the one reason to s- start at the state and not federal level. State level pilots allow faster learning, tighter measurements, and real-world refinement before anything is considered at national scale. And if I learned anything serving in a blended wing, the National Guard can do it much faster than the active duty can. [0:28:34] If this model works well in Georgia, it creates evidence that others, including other states and our federal partners, can learn from. [0:28:46] Okay, this is my last slide. You know, having sat on the other side of funding requests during my time in command and my time in the Pentagon, I'm mindful of the responsibility that comes with making requests like this. So I want to be as clear, specific, and transparent as I can. We're proposing a defined, timed, bound pilot. Specifically, we're asking Georgia to sponsor up to 450 Georgia National Guard members over the next two years. [0:29:18] That allows those members to receive structured documentation, preparation support at no personal cost while they're serving. At the cohort scale, that comes to be, uh, about $1,667 per soldier or airman, or roughly $750,000 over two years. We've self-funded and built this carefully so we could prove the model before seeking public help. We've turned down multiple VCs out of San Francisco because the mission didn't align. [0:29:52] Now we're looking for the right public partner to scale something that's already showing real traction and real value. We think this pilot will allow us continued learning and refinement, and we'll be able to get that $1,667 price down even more and be more accurate. [0:30:15] If results don't justify it, Georgia has a model... If the, if the results do justify it, Georgia has a model that others can scale. If they don't, then we stop. We think this is a measured way to learn what has worked before, before committing or asking anyone to commit to anything larger. We'd welcome the chance to prove and improve this model here in Georgia first. [0:30:41] I'll close by saying thank you for the opportunity to address the committee. And in a way that I've heard many Georgia Guard commanders end any important moment, it's a great day to be a Georgia Peach. I wasn't born one, but I served alongside many, and it would be a privilege to support Georgia once again. Thank you for your time. We're ready to answer any questions you might have. All right. Thank you very much for that presentation. [0:31:12] Uh, questions, uh, Representative Jackson, it's at you, 18. All right. Thank you, Mr. Chairman. Thank you, uh, for the presentation, and thank you for your service. Um, [0:31:27] I can use my Navy commander's voice. Um, couple questions. Thank you, uh, again, for the presentation. Um, do you have any results of the current model that, that's already in place? Are you doing this already? Can you share with this august body some results that you're, you're seeing? You bet. I'll, I'll, I'll, um, I'll s- I'll do a, a broad overview of where we are now and then I'll turn over to, uh, to Marcus, who's our, uh, chief technology officer, who's been doing all the software front-end and back-end design. [0:31:59] Um, we've put about 25 people through the, the, the, uh, through the, the application. Uh, first, um, w- one by one, and then slowly using the application as we continue to build it. This is where we get about 700 total claim issues, of which we already have results back from the VA, of which 92% pass rate, which is a significant higher margin, um, than normal prior to, you know, before anyone were to use, uh, that's, you know, Vetscore. Some of the results that you would, that, the, some of the national results are sometimes less than 50%. [0:32:33] Um, we find that because people are claiming things that are not substantiated in their records. Okay. Uh, as a follow-up, Mr. Chairman. Um, so I know sometimes capacity is built around staffing, and I, you, you gave us, uh, uh, the reasoning why you, you have a small staff and you wanna grow out. Um, can you give us a timeline? Because, you know, I'm looking at the numbers, and I'm a numbers person. I love numbers, but I'm just trying to connect the dots for this, this, this committee in terms of what you're suggesting, uh, in terms of a timeline and what your staffing looks like, so that way, uh, those brave men and women will not experience what they're already experiencing, which is a c- a ca- a clog in the backlog because you don't have the staffing to deal with the unexpected capacity. [0:33:31] That's a good one. Yeah, that's, that's an excellent question. Hi, um, um, as a way of introduction, my name's Marcus Smith. I'm a career technologist, about 25 years experience. I've worked, um, lots of industries, healthcare, uh, startups, media, excuse me, yeah, tech industry. Um, so, so I kind of represent the technical side of this. Uh, to answer your question, like we've been as conservative as possible in the beginning. We're thinking roughly probably, um, one, one staff could handle about 20 service members. And what we're trying to do, though, is leverage, you know, AI and, and things like that, uh, as, as much as possible to kind of like, you know, guide and, and help these staffers be as efficient as they can be, right? [0:34:11] So, so what we're trying to do is find the, this, this alignment between what our internal model systems, and we've worked really hard to keep this kind of like a closed model, right? We, it's, it's very, very important to us that everything's compliant, that sort of thing. But really part of this, this, this setup is to create tools that help facilitate, uh, these support members to do as many as possible. So yeah, so we've kind of like predicated on the amount and, and we tried to be conservative when we think like, you know, what, would they ask for 400 members over two years? [0:34:42] But as to Nelson's point, technology gets better, it's crazy, every week. So we're really trying to get that lower and lower. But again, we have to be so careful because we don't want, you know, we don't, we would never want any, any, any, any sort of like, um, condition to be flagged inappropriately. We always gotta run it by, you know, a human. But, but i- in the same way, we're trying to create these, these efficiencies to make sure that, you know, one human can take 20 for now, but maybe, maybe more and more, and that's how we scale it out, right? [0:35:14] It's really, really based on this kind of like AI model-driven process that we can make it easier for, for these, uh, claims, uh, supporters and adjudicators to, to do as many service members as possible. But your, to your point, yes, we're being very conservative in the beginning. Okay. Thank you. Thank you. Uh, Representative Scott, is that you? Great. Yes. All right. Good afternoon, and thank you, um, for your time. Okay, um, you said you was looking to, um, about 450 National Guard members. [0:35:47] Um, do you, uh, provide assistance for prior military members? Um, uh, people that have been out the Army for years or just, uh, people that have been in, in the Army or in the military. You know, I'm a Army brat, so I always think about the armies. Um, but do you support, do you support, um, active duties, uh, people been on active duty then got out the military, retired? Not people that's for South National Guard, because I know you're looking at 450 National Guard. [0:36:24] Throw in some, uh, prior military. Um, you know, thank you, thank you for the question. Um, to be very honest and transparent, r- right now our software c- we, we, we cannot help, uh, veterans. That's why we've, we've, um... I, I wanna be clear about this. We think the issue is timing. And right now we wanna tackle this one problem at a time. Um, and so we, we have elected to go after currently serving because they have the time left in service to get their records right for the VA. [0:37:00] Those that have already separated or retired, it's, it's, it's even more challenging and harder. Um, but a- as I look everyone here in the face, this is a lifetime passion project for us. I hope to be back here one day and, and say, "Yes, ma'am, we are able to help veterans." Um, there's not, there's not almost a, a day that goes by that we don't travel and talk to somebody, an Uber driver, uh, someone at Home Depot, uh, that shares a story that they're, they need help some way or another. Um, but it's also really important to us to be clear that the issue is not with the VA, it's with missing records, and the best way to fix that is to fix it while they're ser- still serving. [0:37:44] And why we think the Guard is a, is a, is a, again, a better first population for us at the state level, um, is because they're very innovative. Um, and willing to ch- make changes and to, and to, and to try new things quicker than the active duty can, can turn. I, I hope that answered your question as honest as I could. I wish that I could say yes right now, but we need time. [0:38:10] Uh, uh, just to interject, um, what you're describing is our ultimate end game, right? So at the, at the end of what we're trying to build here, I wanna help every service member. I wanna help the, the Vietnam veteran that has things written on paper that you have to scan, you know. We wanna do, we wanna do as wide of umbrella as we can. But to Nelson's point, we really wanna stick to something that, that, that's structured and, and, and, and that, that we know we can, we can prove model-wise, because it's so important to us to get this right, right? [0:38:40] It's so important that no matter what we say we're gonna do, we do it. And, and we wanna make sure that every promise we make to every individual, uh, one-on-one, we figure we can deliver on. And so you're, you're, you're absolutely right, and that is our end game. That's just not our current ask for, for right now in this time. But we'll, you, we, we'd love to re- meet, meet back in a, in a, in a few years and say, "Yes, yes, we're supporting as, you know, we're supporting every veteran we can." [0:39:04] Great. Thank you. Representative Holcomb, is that you? [0:39:08] Thank you, Mr. Chairman. I totally applaud what you're trying to do, and I think probably most people agree with what you're trying to do, and it is a real, um, problem. I'll give you some sort of Shark Tank-type feedback, though, um, which is, one, I disagree with the premise that it's not with the VA, and you may have been just more glib than you intended to be with that, because that, that plays a major role in terms of staffing, the number of people that are able to evaluate the claims, review the claims. [0:39:41] And some of them have a fairly high degree of complexity. And just as background, we don't know each other in terms of introduction. I'm a former Army officer, deployed three times, and then since leaving active duty, I've helped people with claims on a pro bono basis. And it can be very, very frustrating and difficult because you have people who are clearly, clearly wounded, either physically or mentally, and we want them to kind of be at the front of the line. And then you have other people who, let's just say, I was not sympathetic at all to what they were trying to do, and didn't think that they needed to file a claim. [0:40:17] And I would directly tell them, "You're wasting my time, and, you know, really, you might wanna look in the mirror in terms of what you're doing of, like, taxpayer resources." So there, there unfortunately is some gamesmanship that goes with this, and not everybody who truly deserves gets what they need and gets it within, like, a finite or, or a, a good period of time. Um, but with all that, I do have questions besides just the context. One is just a framing, which is agreeing completely with what you're trying to do. [0:40:52] And I also find it disquieting that the answer to a problem for which the country spends a massive amount of public money, both through the Department of Defense and through the VA, is a request for more public money to be used to solve that problem. And with the specific request on, um, on Guards folks, what would be helpful to hear is my understanding, and correct me if I'm wrong, is unless you were deployed multiple times and on active duty, your access to veterans benefits is actually quite limited. [0:41:31] So can you share what scope of, of, of resources you might be looking to help them? Like, I understand active duty folks who get out, and then they have a whole host, um, depending on how long you served and everything else. But their eligibility isn't tied to a question mark, whereas with some Guard members, I think that it is. So can you give us just a little more clarity of, of the scope of how this would be helping? [0:41:59] You, you-- Yes. [0:42:01] Thank you. [0:42:02] Thank you, thank you for, thank you for your comments. Um, and the question, it's, it's really a beautiful one, and cer- the reason that we have-- the reason that we went down this path is because the Guard represents really complex cases, and it really allows us to challenge our s- it really allows us to challenge our software and to challenge our learning loops. You, you know, um, we also know, um, right from personal experience, the 116th has been on many deployments, you know, for many, many years. Um, and the, and the G- Georgia Guard, same, same thing. We still have so many people on active duty in the, both the Army National Guard and the Air National Guard in Georgia that have served in Iraq and Afghanistan, um, that we think that the, the, that it's, it's literally the, the perfect, um, perfect pool of people, um, that will allow us to make the software better, um, and to work through all those issues that you're talking about. [0:42:48] Okay. That makes sense. Um, because one of, one of, one of the things that I was thinking about was if you test modeled this at, say, Fort Stewart or Fort Benning or Moody, is, like, those individuals may or may not have a nexus to Georgia. And so if we're gonna spend Georgia money, there should be some Georgia benefit, one could argue. So I hear what you're saying in terms of, of the Guard space. And then one last question, if I may, Mr. [0:43:18] Chairman, just to make sure that I understand it, is with respect to the cost of $1,167 per person, um, like, roughly how mu- i- is, is there a time spent with them that's correlated to that? Or rather, is it because of AI and the uses of technology, you can sort of scan their paperwork and make sure that they have what they need in there to reflect whatever exposures they may have had? [0:43:49] Like, what, what's-- If you can drill down a little bit more of what is the precise service that you're providing. I get it on the macro level, but if you can get down a little bit more into the micro of, of how it works. [0:44:00] Yes, I will. I'll, I'll try my best, and if it's not, um- If it's not good enough here, I'm happy to send you, um, the, the pricing modeling that w- that we've done. Um, I think what you're asking is, is how do we, how do we come to that, how do we come to that figure and what do they get? And that's, that's also part of the, of, of the point of starting way earlier, right? Right now, when someone has six months, 12 months, 18 months, it's so much work, so much, so many, so many human hours that it, it ... To go through 2,000 records, no matter how good AI is, right, a human still has to do that, and it's very, very expensive to do that. [0:44:36] That's why many law firms charge upwards of three hu- you know, $3,000, $4,000 here in the state of Georgia to go through someone's records, right? And that, I mean, uh, you know that if you've, if you've helped in this space before. Yeah, I don't charge- Yeah ... and I don't think people should out here. Right, right, right. Yes. Yeah. Um, um, but when we start earlier, it's much less work, right? Um, if, if, if they have access to software when they start military service, they don't have any military records to start with. Yeah. All right? And the first time they go to the, you know, to the doctor, you know, once a year or something like that, or once a year, um, it becomes much easier to manage and much more cost-effective to get the price way down. [0:45:15] Does, does, does, um ... I, I just wanna pause there. Does that, that make sense? It does. It's, it's expensive. The closer you are, the more medical records we have to go through, the more expensive it is for our team right now in just time, effort, energy, and money. We, we hope this 450 people, um, is a wa- is a, a, a good sampling of maybe a, you know, an airman or a soldier that's been in for six months, 18 months, and then also someone that has 24 or 26 years. Um, and we think it's a cost average out to, to about that, over two years, which goes back to, I think, your initial question, sir, of, A, you know, we can't put 450 people on this, you know, tomorrow. [0:45:49] Um, but, you know, with, with, with the, uh, you know, the, the, the cleanse lines, we would adjudicate that where, where it made sense, and that's the, that's the average that we came up with based on the 450 over two years with a, a good sample size. Got it. And I'm happy to follow back up, sir. I understand. Yeah. Thank you. [0:46:10] All right. Uh, Representative, Representative Gisler, and then come to you, Ray, and then John. [0:46:17] Uh, hi. Thanks, Chairman. Um, and thank you for the presentation. Um, just kind of looking at this, I, uh, just to give a little introduction of myself, I am actually not a veteran. My father's a veteran, but, um, I've worked in technology for r- sounds like around the same time you have. Um, and I just, just understanding the use case here, i- is the idea that ... I mean, you're, you're talking about someone getting enrolled in this program from day one when they joined the service, and then staying in presumably for the rest of their life. [0:46:49] Is there a recurring cost for that? I mean, the, the presentation sounds like it's a one-time cost, but I just wanna make sure, like, long term, is there a, an investment? [0:46:58] Sure. I'll, I'll, I'll take it, and please clean up if I, uh- [0:47:00] No. Sure [0:47:01] ... um, yes. Right now it's a, it's a, it's a four-year license, right? So it'd be good for four years at that, at that price. [0:47:08] Mm. [0:47:08] And it, you know, our, our, our forecasting models show that price going significantly down, right? Um, over the, you know, the next four years, the next four years, as, um, as we are ingesting these records into the, into the machine. Uh, uh, so, like, [0:47:29] it's difficult to, to give you a precise answer, but I ... we think it's exponentially less money- [0:47:35] Mm [0:47:35] ... as it, as it keeps going, as our software gets better, um, and as w- when they're starting so much earlier. [0:47:42] Mm. [0:47:42] I ... That feels very unsatisfying even saying it, so I, I'm trying to, uh, like, from your perspective. Um, I just, I don't want to throw out a dollar number that's not valid, but, but it's, it's, it's, it's half or less, and then just gets, gets better and better for the taxpayer as we go. And if I may- [0:47:59] Mm-hmm [0:47:59] ... um, to address your, your great point about, you know, why are we paying more money now, uh, for a problem that, that we've already spending a lot of money on, uh, w- we believe that this is gonna save the VA and the American taxpayer hundreds of millions of dollars, and we're very close to having that model, uh, you know, ready. And, uh, Stanford is helping with that. Um, we're talking, because you're, you're ... We're talking in, in actual time spent re-adjudicating claims. We're talking time spent of, uh, there's no reason, uh, to p- um, to, um, to refile all of the extra paperwork, all the extra time, effort, energy that the VA has to spend, that the reviewers have to spend. [0:48:39] That decreases. So not only do we believe that we're gonna get our cost way down here, but we think the cost w- it's, it's scaled federally, is hundreds of millions of dollars to the American taxpayer. [0:48:52] Yeah. I, I, I can understand that, and I can see how there's economies of scale that play, you know, the further you go with it. And it's a new product. I've, I've worked in startups. [0:48:59] Mm-hmm. Yeah. [0:48:59] I understand that. Um, so that, just to be clear, though, that 1,667, that would be for a four-year license per member? [0:49:07] Yes, sir. [0:49:07] Or is that per year? [0:49:08] No, that's a, that's four years. That's 450 people for f- for four years. [0:49:11] For four years. Okay. [0:49:13] All right. Um- [0:49:15] Uh, do you have ... Uh, the 92%, you know, first pass, you know, rate is impressive, quite honestly. Um, do you ... Uh, it should be easy to kind of turn that into an ROI projection for the state. If we were to invest that money, what kind of ROI could we see? Uh, is that something we can get at some point? [0:49:34] We'd love to do that. Thank you. [0:49:35] Absolutely. [0:49:36] Thank you. [0:49:37] Thank you. [0:49:38] Thank you. Um, so that, that is an excellent point. Uh, it's pr- one of the questions that I, that I had is that it would be great to understand the economic benefit to the state by Georgia residents receiving those benefits, right? Uh, it's, it's, it's something I know that Commissioner Ross's, um, uh, agency, uh, you know, is able to produce in terms of benefits into the, into the system essentially, uh, that, that helps to, um- [0:50:04] Again, spell out that return on investment. All right, Representative Martinez. Thank you so much for your presentation. I see here that you got some sponsors, foundations, uh, nonprofits and so on and so forth. Do you have a team, like a promotional team, a grant writing team, that would solicit these funds for your organization? Um, we do. A s- a small, a small team. We have a, a separate nonprofit, um, that, that works with us, that, um... So that one answers your question first, y- yes. [0:50:35] Um, and then Stanford University has supplied us a research assistant to write grants. Yes, sir. Thank you. [0:50:44] All right. Uh, Representative Scott, you had another question? Yes. Um, um, how would you, uh, recruit these four hundred and fifty, um, National Guards? Uh, what, what is the process, um, to get them on board with your program? Um, talk to their commanders. Um, we have a, um, um... [0:51:12] We have a this best suits or is best used for here, but, uh, as a f- former commander, I, I know that, um, um, I would not want to decide for a commander, you know, who gets to, who gets to be put in or not. We'd like to give the guidelines and allow, allow the commanders, uh, to select, uh, in the units. We're looking to, to do a split between the, the Guard, the, the National Guard, um, and the Air National Guard. Um, and that al- that also, to the question of, of why this is so, um, why we think this is such great data, um, is because we have, uh, flyers, um, in the, uh, 116th Air Control Wing up in Savannah, um, and they often present different, um, different health issues than those that have fought on the ground. [0:52:01] All right. I don't see any other questions. You, you... One more. Representative Jackson, go right ahead. And, uh, we just-- if you would, kind of make it quick. We just got... [0:52:12] Thank you, Mr. Chairman. I, I was using AI, Mr. Chairman, and AI came up with over a thousand different combinations. And I know on page nine here, uh, you, you, you have a vet score approach around health issues. And so I'm just-- a- and this goes back to, to my colleague's question as well. I'm just trying to figure out, rationalize in my head, every service member, especially the women, right? [0:52:47] My, my first wife served in the military for 27 years. Her health issues were different because every birth that she had, um, with our children was cesarean. And so I'm just trying to quantify your approach with the various different combinations that AI is saying that a service member can experience, and a lot of them are with our brave women who are in uniform. [0:53:18] And I'm just trying to make sure that there are no gaps here, because, uh, going, going back to, to my dear friend, um, the VA is already a complicated system, and I just wanna make sure that we're not adding another layer. Yes, I, I appreciate everything that you said and you p- present it to this august body. I just wanna make sure, though, in the end, the brave men and women are getting the services that they need. How do we resolve that through the use of this vet score? [0:53:52] Yeah. Excellent, excellent question. So probably people are, right now, veterans, right, they're probably uploading records into ChatGPT or what- whatever. You know, it's, it's happening, right? AI's already, like, kind of, y- you know, worked its way into our daily lives. And that's a lot of this caution and conservatism we're having, right? We have-- We worked really hard to set up, like, you know, very HIPAA compliant Amazon, you know, uh, a setup with them. We have a private LLM. We have our own training models. And we, we, we look at, like, every single record that comes in right now, and that's why we're starting small, is we, we wanna do that contextual analysis, right? [0:54:29] It's very, very important to us not to say, like, "Hey, we wanna splice something out and see what, like, a, a cursor or a ChatGPT or whatever, whatever it has to say, or Gemini, what- whatever that has to say about it and all these permutations." We're trying to be as, um, a, a slow approach and, and, and to work closely with people that, that are experts in, in this kind of industry and say, "All right. Why is this a positive? Why is this a negative? Is this a false positive? Why? Can we, can we add a rule around that?" So a lot of the work we've been doing in the past is kind of building up that rule set, right? [0:55:01] It's-- And that's gonna become, I think, you know, year one, year two, uh, through this pilot, that's gonna become one of the most important things to get out of this. Sure, we're, we're helping, uh, uh, uh, the, the set, and, and, and I hope, and I think we're gonna do a really good job at it, but it's also getting this, this kinda, like, I don't wanna say proprietary, but getting this vet score rule set that we can say that, like, "Hey, put it through our system and feel, you know, like we have you. We know exactly what the context is here. We know exactly, like, you know, we've seen all these other records that we've had a human really, really look at and, and, and kinda help us decide, you know, w- what, you know, is it family history versus, like, this type of, type of birth you might've had or, or whatever." There's so many of these, like, you know, I'd say, like, facets, right, to, to how, uh, how these, um, things are, they're analyzed, that that's for us, like, the most important thing, right, is really getting that right. [0:55:54] And then also representing tools that say, "Hey, if you're, if you're helping the veteran, if you're the c- y- you're the caseworker or whatever, we're just gonna put the information in fr- in front of you. This is the best that we can discern. You make the decision," right? We're not gonna have technology make any of these decisions. It's gonna be supplemental. Now, when you look at, like, what we're trying to do year over year, though, is make that even easier, make that even, even, even more empowering for, for these, these caseworkers to move faster and get a really good, um, you know, e- e- enablement into what these, you know, records pertain. [0:56:26] Does that-- I hope that [0:56:28] That helps a little. [0:56:28] Yeah. [0:56:29] I guess, Mr. Chairman, I mean, because in the end, when you think about, it's still gotta... The, the, the, the, the record still must go to the VA, right? It's still got, it's still gotta go to the VA. It, I mean, because they're not the approval authority here. I mean, they're the, they're assisting the service member to make sure that the I's are dotted and the T's are crossed. I just wanna make sure that all the service members in Georgia are, are getting exactly what they need, and I, I'm, that, I guess that's where I'm, I'm perplexed at. [0:57:12] Yeah. Well, I, I do appreciate that comment. We'll, we'll close out with this. Uh, fir- you know, thank y'all so much for, for what you're doing, what you've come up with. You're, you're certainly serving a noble purpose, um, and, and you know, I, I think it's wonderful, uh, on, on that side. Uh, I, I share your frustration, Representative Jackson, in that we have, we have, uh, uh... There's no other way to say it. We are failing our service members. [0:57:38] Mm-hmm. [0:57:38] You know? We, we, um, you know, we are, um, asking folks that have volunteered to, to sacrifice everything, and we are allowing them to transition out of the service with incomplete records, um, with issues undocumented, untreated, uh, unreported. And then we're putting them into a system, uh, of the VA that is broken, that requires an entire industry surrounding it to help a, a n- individual navigate that system, right? [0:58:11] Um, I've, you know, my ex- my, I grew up in the Army, um, and, and the phrase that was instilled in me is taking care of soldiers. You know, in, you know, input whatever branch of service, sailor, airmen, Marines, but that's what I was taught, that everything you do is to take care of those soldiers. And I think that, and, and this, gentlemen, this is not a reflection on, on, on y'all. You guys are trying to solve this problem that we, uh, have allowed to, to be created, uh, in, in our government, and it's, and it's, it just boggles me and frustrates me to no end. [0:58:44] And it's much like dealing with the IRS. We've got an entire industry of, uh, of CPAs and, and accountants to tell us how to pay, uh, taxes to the government that created that system, that no one can, no one can tell you and pick up the book and, and navigate it themselves. So, um, what y'all are doing is certainly noble. I wish that, that, um, you did not have to come up with this kind of technology. Um, and so while I certainly support the effort, my, my frustration is more on, on what led us to the, to this problem that we're having to solve. [0:59:21] And, and I think in so many ways that we, we need to, to press the VA to fix. Um, we need to empower, continue to empower our own Department of Veteran Services to address the issues that they've identified and, and do what we can to take care of our service members. So, uh, appreciate so much the presentation. Uh, I'm, I'm, I'm amazed at the, the technology and the, the ideas that y'all come up with. And, and, uh, for myself, I am certainly supportive and would look forward to, uh, hopefully seeing you guys again soon. [0:59:53] All right? Uh, if there's no other business before the committee, we will stand adjourned. Thank you. Thank you so much, everyone. [0:59:59] Thank you.