← Embracing Digital Transformation
Embracing Digital Transformation
#352 How AI Is Fighting Fraud and Expanding Access to Public Services
00:30:19
Loading summary
Transcript69 lines
- [00:01]
A
Dish has been connecting communities like yours for the last 45 years, providing the TV you love at a price you can trust. Watch live sports news and the latest movies, plus your favorite streaming apps, all in one place. Switch to Dish today and lock in the lowest price in satellite TV, starting at $89.99 a month with our two year price guarantee. Call 888-add-d Dish or visit dish.com today.
- [00:30]
B
It can help kind of vet that data to see if it fits the eligibility criteria or not. And then it can put a really, really clear, tight workflow in front of an eligibility worker to either approve that person, request more information, or deny them for the program.
- [00:49]
A
Welcome to Embracing Digital Transformation, where we explore how people process, policy and technology drive effective changes. This is Dr. Darren, Chief Enterprise architect, educator, author, and most importantly, your host on this episode. Fighting public service fraud and increasing accessibility with AI. Today's guest Aid Kit CEO Brittany Christensen. Hey, Brittany. Welcome to the show.
- [01:23]
B
Thanks, Darren. It's a pleasure to be here.
- [01:26]
A
Hey, everyone that listens. Well, let's talk about the topic. You have a fascinating background, which we'll get to in a moment, but you're making some fundamental transformations in an industry that's overlooked a lot of times. And do we call it an industry? All right, I'm teeing that up. We'll tease it a little bit. But before we do that, everyone that knows listens to my show, that I only have superheroes on the show, and every superhero has a background story, an origin story. So, Brittany, what's your origin story?
- [02:03]
B
Thanks for asking, Darren. So I grew up in Northern California, and I come from a big family. I've got five siblings. And I, yeah, I just, I grew up with a, with a major love for nature, and as you know, quite the nerd. My older brother and I especially, always, always, always, since we were children, have stayed up late talking about the universe and physics and what we want to change and why most people's lives are boring, but ours are not going to be boring when we grow up.
- [02:35]
A
That's so awesome.
- [02:38]
B
And anyhow, you know, just had a lot of intellectual stimulation and passion from a very young age. And I ended up going to college in Montana at Carroll College, a private Catholic liberal arts school with a really cool outdoor program, a great math program, and I got my undergrad degree in applied mathematics, which was a ton of fun. I've always just appreciated the kind of sense that math makes. It just is very calming for me. But after undergrad, I went to grad school for a short time And I was focused on epigenetics. I went to McGill for a biostatistics program. And I found that things were getting a little bit too theoretical for my liking. So I went back to the drawing board. And what I intended to be just a summer away ended up being a seven and a half year career farming. So I went from a math nerd to an organic vegetable farmer.
- [03:42]
A
No, wait, where did you set that farm up? Back in Northern California.
- [03:47]
B
So this, this time I, you know, I've gone over to McGill, which is in Montreal, Canada.
- [03:51]
A
Yeah, yeah, I know McGill actually really well. Yeah.
- [03:54]
B
Oh, wonderful. Yeah. So the farm was in northern New York, which is about just a couple hours south of, of the university. So a friend of mine from college who ended up being my husband and I and our best friends ran this organic vegetable farm, the two of them. It was their brainchild first, so they had already started the farm. And when I left grad school, I came down to help them run it while I went back to the drawing board on my own career. But I just fell in love with it. You know, I fell in love with kind of direct, immediate, tangible impact and, you know, getting out of my own head and getting into the physical space a little bit more. And it also was really connected to environmental justice. So we started at the tiny little farm, three acres on an old hay field. And it ended up growing into 90 acres and feeding thousands of families across the state of New York, including going all the way down to the green market in New York City. And it just taught me a lot about what it looks like to help people. And also it showed me a lot about who kind of falls through the cracks of a lot of our existing systems and got me interested in food justice. So that kind of path from farming to food justice put me in the seat of leading a nonprofit called ADK Action. And that was another six year journey. A lot of really fun work there, just kind of helping get good organic food into the hands of low income people. And we did other community development projects as well. And somewhere during the pandemic, I kind of hit a wall with my own inspiration. I kind of ran out of gas, if you will. And what I realized was all of this great work we were doing could be completely undercut by the next policy change.
- [05:47]
A
Oh yeah, yeah, totally could, yeah.
- [05:51]
B
And you know, anything you do to work on the food system, at the end of the day, the economic system is just going to trump that. So it's really about how do we pay people fairly for their labor in the fields. How do we make food affordable? Excuse me? How do we make food affordable for families all across the country, no matter their background? And how do we help people in a way that's not stigmatizing? How do we make it feel okay and feel good even to ask for help when you need it? And all of those big questions are what landed me at aid kit.
- [06:28]
A
You're tackling some pretty big problems there, Britney. Those are big. Those are big issues.
- [06:34]
B
So I guess I'm drawn to the big issues.
- [06:37]
A
You're drawn to the big issues. Does it scare you a little bit to go after such big issues?
- [06:43]
B
The only thing that scares me is running out of time before I can solve any of them.
- [06:48]
A
Okay, all right. So you're a big problem solver. So you come. If you're in Northern California, then you probably come. Were you in the. In the Sacramento Valley? Is that where you were at?
- [07:00]
B
Oh, so that's funny. That's where my mom is from, but we actually grew up even much further north, like Crescent City, Gaskey area.
- [07:08]
A
Okay, all right, so you're up on the coast then. Because I was going to say the Central Valley of California is the breadbasket of the world. So you had farming. Even up in Crescent City, you still have this farming agro kind of feel to it. So maybe. Maybe that was in your. In your blood.
- [07:29]
B
Yeah, I think so. I've always been drawn to it. There's nowhere more peaceful than the garden. And yet here I go. I got myself into all kinds of trouble. Now I'm leading a tech company. So that's not as close to the garden as I used to be.
- [07:42]
A
So let's talk a little bit. Let's tease that out a little bit. Because you're running a tech company in the food industry. This sounds like one of the last industries I would think that would be using technology. But that's not the case as I've dug in a little bit. In fact, food production is all techs, all over food production in huge ways. Can you kind of explain to our audience here what's going on in that industry as far as technology and where you see the future?
- [08:17]
B
Yeah. So actually, the work I'm doing now at Aid Kit is not food systems work, it's economic justice work. So the tools. Yeah, so, I mean, there's definitely a connection, though, because what we do is we deliver direct cash assistance to families all over the country. Actually, we just crested the $420 million mark of dollars out the door into people's pockets. So we're you know that money is definitely helping them eat.
- [08:53]
A
Yeah, yeah. When you say into people's pockets, not into bureaucrats pockets, not into. You're dealing directly with the public, right? With the people in need. Right?
- [09:03]
B
Yeah. Our business model, we actually partner with governments and nonprofits and, and then we help them directly deliver the cash. And so it's a two sided equation. If you're talking about industry, it's probably more like a B2B2C motion. And our mission is to deliver aid with dignity. And the way we do that is we work with folks in government who have innovative programs that are seeking to reach vulnerable communities. So maybe it's a cash assistance program for individuals who are unhoused or, or maybe it's a cash assistance program for refugee resettlement for instance, or guaranteed income. We come in as the technology partner and we bring a human centered design lens to putting together the infrastructure that's going to help government reach these communities at scale and verify their eligibility and ultimately pay them out the benefit that they qualify for.
- [10:03]
A
Okay, let talk to me a little bit about that. What role does technology play in this? Because I would think, come on, there's already systems out there that are supposed to be taking care of this, but what I'm hearing is there's a gap, seems to be a gap. So what gap are you filling there? What is that technology gap that you guys are filling out?
- [10:27]
B
Yeah. So for decades and decades government has purchased basically consultant built bespoke systems that are, that are created for the specific program. Sometimes the government even wants to own the code and then do M and O maintenance and operations contracts on top of that. I know you're a public sector guy in your background intel, so you know this better than anyone. We, we come in with a product that's more like a SaaS platform, but in a really cool way. Because the problem with SaaS is especially for this sector, it's normally one size fits all. And so SaaS products are trying to solve for that middle of the bell curve. For most users 8 kit's doing the complete opposite. We're solving for the low income, vulnerable, very difficult, hard to reach tail of that bell curve. And so that means that the systems that we build need to be super customizable and they need to be able to be configure it on a community by community basis. So we've basically taken the super bespoke high touch consulting model and we've productized it in a way that leads to better outcomes for the communities that are being served And a much lower cost, faster to configure technology solution for our government partners.
- [11:42]
A
Okay. I mean, that's a tall order. I know because my wife and I have done some volunteering for the IRC here in Sacramento area and we got a whole lot of Afghan refugees. And then right after that the Ukrainian refugees came in. And I did notice that it was hard for especially the Afghans when they first got here because there wasn't a landing pad for them. There wasn't a community already here for them where the Ukrainians. We already had a good Ukrainian immigration here. So finding the services and things were very difficult for, for these people. And when they would ask me questions, I had no idea what to tell them. So what you guys have come up with is, is something that helps those very specific or specialized types of parts of our society that the need that. That are out of the ordinary, is that best way to put it.
- [12:43]
B
Absolutely. And it's just like universal accessibility and building code. Right. If you put in a ramp, it's not just folks in wheelchairs who can use the ramp. Moms with strollers can use it. And in fact, anyone with legs can also come right up that ramp. So we build for the hardest to reach communities. And what that enables us to do is reach everybody.
- [13:03]
A
Gotcha. All right, this is starting to make a little more sense to me. All right, so the big question is, is what impact is AI having on this? I would think AI would have a huge impact on this, on this industry and make it even easier.
- [13:18]
B
Yeah, absolutely. So, you know, one of the things. So just to kind of explain the lay of the land of what it is that we actually do, we have, you know, these front end applications. And I'm going to use that IRC example with Afghan refugees because actually IRC is one of our customers and we, what we do is we have a front end interface for people to be able to apply for or enroll in benefits. And that interface, like on a website, Right?
- [13:44]
A
That'd be like a website. Okay, gotcha.
- [13:46]
B
Exactly. We use a progressive web app. So it's in browser, it's multilingual, we can support up to 120 different languages. But we go even a step further and we hire native speakers and oftentimes people that actually have been in their situation. So we did hire Afghan refugees to be translators for our IRC program. And they help us design an interface that's culturally responsive as well as accurate with the translations. And then when folks come and enroll, they get a cell phone as part of their as part of the program. And they're going to get a text message that they can read, that they can understand, that's written for them by someone who knows them, knows their culture. And then they can enroll and they can select how they want to get paid. And they can get text messages again in their native language, giving them access to resources, helping them identify community groups that they can meet with, that sort of thing. But prior to enrollment. And it looks different in every program, but in a lot of our programs there's this big enrollment step where people have to demonstrate that they're eligible for the program. And every program has different eligibility rules. So that's one of the big use cases for AI is, you know, typically you're talking about filling out really long, complex forms with a lot of documents being uploaded, just like we're all used to every year when we re enroll in health insurance, for example.
- [15:10]
A
Right. Yeah, I know, yeah.
- [15:13]
B
But AI can connect those dots a little bit more easily for everyone. And it can kind of help, especially on the back end, when you have the data that an applicant has put in, it can help kind of vet that data to see that if it fits the eligibility criteria or not. And then it can put really, really clear, tight workflow in front of an eligibility worker to either approve that person, request more information, or deny them for the program.
- [15:48]
A
I love what you're doing here because, I mean, we, we see the news, it's, it's all over the place. There's so much fraud in a lot of these programs. And the first reaction that a lot of people have is just cancel the programs because there's too much fraud and there's too many people out there helping, helping these people with fraud. But what you guys had come up with, it sounds to me could help prevent a lot of that fraud from happening by having AI actually, you know, kind of intercede and say, hey, they do not qualify for these, don't try and force fit them in, because it doesn't make sense. Right. Is that one of the benefits that also that you can get from this sort of thing?
- [16:30]
B
Absolutely. Yeah. And as you know, as we've talked about, we have a big focus on accessibility, making sure that people really have a good path in, even if they face barriers to applying. But there's been for a really long time, especially in government, especially politically, kind of a false trade off that everyone's been fed, that you have to choose between accessibility and fraud prevention. And with modern tools, that's absolutely not the case. You can have your cake. You can Eat it too. You can make sure the people who need help are getting it in spades and that bad actors have no chance.
- [17:01]
A
Well, and in fact, it's kind of an oxymoron when you think about it. Right? Because the bureaucracy that they put in place is supposed to dissuade people that don't qualify to getting involved. But what happens is the fraudsters, they spend the time figuring out how to work the systems, and then once they get that, that methodology down, they just crank. They just crank it out. Right? And it's. And the only people being hurt are the ones that really need the help because they can't jump through all the hoops that were established that people thought would filter out those that weren't in real need. Is that kind of the.
- [17:40]
B
That's absolutely accurate. Yeah. And I think what a lot of the. A lot of people don't see is the huge collateral damage that has on the back end on the state agencies, county agencies, and the eligibility staff. Because not only do the people applying have to jump through all of those hoops, but the people that are qualifying them, they have to jump through those hoops on the back end. But they've got four or five hundred cases lined up on their desk and they're working with sometimes packets with handwriting that have been dropped off in person, and emails and text messages. Oh, it's a nightmare. So that's one of the things. Our core value is putting recipients first. So all of our interfaces, we put all of our initial focus on building the best possible interfaces for people seeking help. And then our second most important customer is our caseworkers. And so we really try to make their lives easier, and we try to blow them away with how much more organized and sane their workflows can be.
- [18:43]
A
Are you seeing them take advantage of more AI in those workflows? And is that something your platform is enabling? Yeah, I mean, it's fairly new still, right? I mean, agentic flows are pretty new still. But do you see, do you see you guys progressing even further in that space?
- [19:03]
B
Yeah, we do. We actually, we just launched a product called Luna. It's a chatbot to help people apply for their benefits. One of the ways that people get stuck is that as they're filling out a form, the way that a question is asked, they might not have a clear or binary answer. And so they need a little guidance along the way. So we have kind of an assistant, a co pilot, if you will, that helps with the application process. And then we're now extending that to the back end. So that when the eligibility worker is trying to toggle between the eligibility rules, the policy intent and some strange edge cases that have come across their desk, they have support in kind of seeing all those different decision, you know, factors that would influence their decision coming together in one place, in real time with real guidance. So, you know, there's a ton of opportunity here with AI. At the same time, we're really grateful that we're, you know, we're five years in, so we have, we have a lot of data that predates AI. So that's really, really helpful because it helps us to. Yeah, it helps us have the data sets and we use a lot of synthetic data sets as well. But you know, it gives us the opportunity to de, identify some of the historic data sets that we have and then use it in such a way as to be able to make more accurate decisions and to provide guidance to the real professionals, which are the caseworkers. Because at the end of the day we don't want these, these can be really, really influential programs in people's lives. And so while the AI is getting better, we don't want it to have to be accountable for whether or not somebody gets life saving support support. So that's always still on the caseworker.
- [20:42]
A
You know what, I love that you said that because that accountability is extremely important. Ultimately the caseworker is still accountable for, for this sort of stuff. So. But if the AI can give that caseworker better information, more reliable information, then that caseworker will be able to handle and make many more decisions much quicker, which should alleviate the caseloads dramatically.
- [21:12]
B
Exactly. Yeah. The last thing the caseworker needs to do is to be toggling between 14 different screens and interfaces. So if they can have all of that context right there in front of them when they're making the decision, they're happy, we're happy, and there's very little if any increased risk.
- [21:29]
A
Where do you see the future going with this, with this industry? I hate to call it an industry, but it is an industry. Right. Where do you, where do you see AI's role in this in the future? And do you see major shifts in, in this industry because of artificial intelligence?
- [21:51]
B
Yeah, I think it's a great question. I mean, absolutely. I think that there's a shift that's happening right now towards more product led solutions instead of reinventing the wheel with bespoke code every single time. I think another thing that's, that's really shifting right now is there's been a lot of policy change and just a lot of. Yeah, you want to. It is an industry, the aid industry, if you will. We're talking about nonprofits, we're talking about governments, we're talking about for profit consulting firms, we're talking about govtech agencies like Aid Kit. So if you look at just the last year and a quarter alone, we saw the dismantling of USAID, we saw the passage of HR1, which is some of the largest benefit cuts in 60 years in our country. And every time these kinds of things happen, it's a shift of resources. Right. So there's been a lot of money that's now being taken out of the social sector and the aid sector, and that's forcing the agencies and nonprofits that deliver aid to do more with less. And so they need to be more efficient, they need more automated workflows. Their staff are overwhelmed. And so, yes, there's huge shifts right now, and there's a lot of efficiency to be gained with modern tools and especially modern tools that can responsibly incorporate AI.
- [23:21]
A
All right, so. All right, so it's kind of a. What I, I called a stimulated hyper transformation. Right. Just like Covid did that to the IT market. Right. All of a sudden we all had to start working overnight. So, you know, in, in the aid industry probably needed a good, a good change. Right. Probably not as dramatic as we, as we got, but hopefully there's, there's a lot of good out of here around efficiency and more equity among how things are distributed. Because I think we're just now unpacking what has actually been going on for, like you said, like 60 years. It's, it's just piled up over time with no checks, no efficiency gains. Everything just keeps, the bureaucracy keeps growing, right?
- [24:17]
B
Well, I don't know if I'd say that exactly. I think that there's. I think that it's a really difficult. I'll put it this way, there's a lot of people who devote their lives to making aid delivery be efficient and equitable already in there. And there have been for a really long time. These things are so controlled by political headwinds and tailwinds that progress that's made is frequently erased at frequent intervals, if you will. So that's one of the challenges. I wouldn't say that it's like, you know, I think it would be a miscategorization to say that the massive divestment from aid of the last year and a quarter has been like a really positive shift. I think it's just been Another in a series of significant challenges that aid workers have faced. But I think that if you're, if we're looking for silver linings. Certainly I'm looking for silver linings.
- [25:12]
A
Yeah.
- [25:12]
B
Becoming more efficient is good. And, and we just, you know, we try to do everything we can to, to help these workers who really are in it for the right reasons. I, you know, I noticed your hesitancy to call it an industry and I understand that, you know, and in, when you make helping people in industry, there's kind of an inherent conflict there. Or it can feel that way.
- [25:32]
A
It feels like a conflict. Right? Yeah.
- [25:35]
B
So.
- [25:35]
A
All right, so here, here's my. A big question. Do you see any profound changes in, in policy that could come about because of more or better efficiency or better tools? Because right now. And that's where, that's where I'm kind of looking to.
- [25:54]
B
Yeah.
- [25:54]
A
Are there any profound process changes that we could see or policy changes that could happen that could make it even more effective? Even more so than efficient, but effective?
- [26:07]
B
Yes. I love that. It's a great question. So one of the things that we talk about and think about a lot at Aid Kit is disruption. Aid Kit was built during disruption. You know, we were born out of the pandemic. So from the very beginning we've been in go mode. And one of the things that's kind of inherent in the industry that we work in is slow timelines. So you see these state systems get purchased and it's like a seven year implementation and it's $150 million contract. It is crazy. Yeah. And in a world where things are relatively stable, you can limp that along. Right. But as you have more and more disruptions, what you find is that that model breaks. And that's what it has done really, since the pandemic. It's broken again and again and again. And at this point it's just obviously untenable. But, but let's think about during the pandemic. Pandemic, unemployment insurance, or pua was put out to the public and it basically blew the top off the eligibility rules for unemployment. People could get, they could qualify more easily and they were getting larger checks and in many ways it was a fraud.
- [27:15]
A
Went up through the roof. I mean, they're just uncovering that rampant. Oh yeah, they're uncovering that in California. $20 billion.
- [27:22]
B
Yeah.
- [27:23]
A
That's craziness.
- [27:25]
B
It is craziness. And the reason that that happened, Darren, is because the systems that were in place, they couldn't, they couldn't change the rules. Overnight, they didn't have the technology, was not flexible. Yeah. So that's really the thing that we're working on is aid infrastructure that adapts faster than disruption can break it.
- [27:43]
A
That's what I'm looking for. Right? That's the golden ticket right there, right?
- [27:48]
B
Yeah, exactly. Because, you know, the thing is, like, we're gonna, the disruptions are only going to continue. AI is an incredible technology, and we already see it changing the nature of the workforce. We have to have new policies that adapt to allow communities to adapt to the impacts of AI and the labor market, for instance. So if we try to take those same unemployment insurance systems and once again, we try to get them to operate under new rules, they're going to be slow to change, they're going to be fragile, they're going to break. So we need systems in place that are adaptable, where you can change the eligibility rules overnight and you can trust that the system's going to accurately enforce that and control against that fraud. And that's what we're doing and make
- [28:29]
A
sure the right people, I hate to call it the right people. The people that need qualified. Yeah, the qualified people are getting the help that they need. Yeah, I, I kind of wish our systems were a lower, a more. I don't know what the right efficient, effective, whatever the, the case may be, but they aren't. And we, we know that, but they are completely out of control right now. So hopefully you guys can come in and just help change the world because, Brittany, I think you're on to something here.
- [29:06]
B
Thank you. We're sure trying.
- [29:08]
A
If people want to find out more about what you guys do, where do they go? Where do they find out more?
- [29:16]
B
AidKit.com is the best place they can also, you know, come on LinkedIn. We have a pretty active LinkedIn presence. Aid K I T and I'm also more than happy to have conversations so you can reach out to me on LinkedIn and send me a direct message. But, you know, we're active across 27 states now. We've delivered $420 million to folks individually. We're working at the statewide level in three different states, lots of different counties, lots of different nonprofits. We do emergency and disaster response in addition to more of these public benefits that we're talking about. And we're, we're just seeing that there's a huge appetite to do better. And we're excited about that wave and happy to talk to anyone who wants to get on board.
- [30:03]
A
I, I think this is awesome. Brittany, thanks for coming. On the show. I sure appreciate your time.
- [30:07]
B
Hey, thanks for having me. Darren.
- [30:15]
A
Thanks for listening to Embracing Digital Transformation. If you enjoyed today's conversation, give us five stars on your favorite podcasting app or on YouTube. It really helps others discover the show. If you want to go deeper, join our extra exclusive community@patreon.com embracingdigital where we share bonus content and you can always connect with other change makers like yourself. You can always find more resources@embracingdigital.org until next time, keep Embracing the Digital Transformation.