Back to podcast

Jingle

Granting the Future from EY.

Tim Smith

Hello and welcome to Granting the Future, the podcast series from EY for grant and public fund managers around the world. 

I'm Tim Smith, your host, and each episode we'll be sharing expert insight about transforming the grants process to overcome challenges and meet citizens' needs.

Now, this time, we're looking at the role of technology in designing programs for purpose. Joining me to discuss this from the U.S. Department of Health and Human Services is Andrea Sampanis, Solutions and Services Lead, Grants Quality Service Management Office. Andrea, hello. Welcome to the podcast.

Andrea Sampanis

Hello. Thank you so much for having me.

Smith

And also with us is Amy Fenstermacher, EY Americas Grants and Relief Funds Management Leader. Amy, hello, welcome to you.

Amy Fenstermacher

Hi, Tim. Thanks for having me. 

Smith

Well, Andrea, let me start with you. Just tell us, first of all, if you will, please, a bit about your role and your main areas of focus.

Sampanis

Of course. So, in the United States Federal Government, we give a lot of money in grants. Actually, every year we give out US$1 trillion in grants. It's the number one way the US Federal Government spends money. And we do that because we need help from all of our partners, states, local governments, universities, tribal organizations to help us do the work, the critical work of the federal government. So, it's a lot of dollars going out the door and a lot of technology needed to help us manage and track all of those dollars. 

At what we call Grants QSMO, Quality Service Management Office, which is sort of a confusing name, I'm in charge of creating a marketplace, sort of a shared services shopping experience where agencies can go to identify solutions, the IT solutions, to manage the grants they have. And also, if that wasn't hard enough, we also try to work with all of those recipients that get the money to make sure we're understanding their experience and their challenges and bring ways to mitigate the challenges through technology, where we can find those solutions.

Smith

Well, you mentioned technology there. Can you give us a sense of the developments, both in the US and globally, regarding the use of technology in grants management? 

Sampanis

Yes, I might say, a critical piece to using technology and sharing technology — what we've learned from other countries too — is the need to first define common standards. You have to sort of do the work in a similar way so that you can actually share the technology. If not, you'll spend a lot of money building your own to do it the exact same way you used to do it 20 years ago. I'd say, overall, we're seeing a major change, not just in the federal government in the United States, also at the state level, but definitely abroad, in first creating those standards and then seeing the value of sharing technology and sharing the information. 

Because once you have the standards in the same way, you can share that information so much better. And as technology is growing, it's making it easier and easier for us to share that. 

Smith

Well, let me bring Amy into the proceedings now. Amy, what are you seeing then as the key themes and success factors for implementing technology to transform the grants management process?

Fenstermacher

The first is thinking about the users from the very beginning. And when I say users, I'm not just talking about the people who point and click in the system, but I really think about, and I think many of us do, I think about grants as a system of systems. So you have people who are supporting from a call center perspective; you have teammates and implementation partners who are providing technical assistance and helping grantees to actually execute on the expectations and compliance requirements of the grants; and thinking about ways to use technology to make that entire process simpler and more efficient and cleaner is sort of a critical piece of the puzzle, because otherwise, you're using technology to automate processes that were suboptimal, to be gracious, in the first place. 

The other piece that really sticks out to me is that automation can also be challenging when we don't think about the repercussions of the automation choices we make on that system of systems. So, when we design automation functions, we're really thinking about how the automation operates in context. What we've seen over and over again is that our clients, or our government counterparts, are trying to use a system that ultimately sort of automates them into a corner and then they get stuck, and they end up actually taking that process offline and doing it the old-style way in Excel because the system is not working the way they need it to. 

And this becomes particularly clear when we think about novel types of grants or novel environments in which we're making grants. So, COVID-19 made this starkly apparent to us that we had to think of automation within the context of a broader ecosystem.

Smith

Andrea, we're in the midst of an AI revolution. We hear about it, we read about it all the time. How is AI contributing to the design of programs for purpose and making the end-user experience easier?

Sampanis

I think a big part is the end-to-end experience, and our goal, it's a very lofty goal, of the idea of “tell me once”, right, I'm going to collect your data once, and hopefully I don't have to keep collecting it because I'm going to share it. But I want to take a step back. AI, people think of it as big robots and really into the future. But truthfully, we have a lot of it working now in our systems. But I think it's important just to share in the United States what's happening. We want an end-to-end experience. We want that recipient to not feel like they're going in and out of multiple systems, which I guarantee you they do now. 

We are not fully connected. So, the end to end, for us, has never been a goal of one system that you go in and every single person uses it. It's just, quite frankly, we've tried that twice in the federal government in the early 2000s and, again, 2010s, and we have failed at it. So, this time we decided, hey, we've already got some centralization around some of our systems. SAM.gov, you go in to register. Grants.gov, you go in to find a grant. Those are different systems and they're going to continue to be different systems because if I merged them all together, it's too much, it falls under its own weight. 

But what we can do is try to share the same login across and share through AI and just APIs, just connect to the systems so that the data will flow, and you don't have to keep telling me. And, as Amy said, it's a huge benefit to fraud and you don't need to rekey things and there's a lot of human error in that. So, we want to create that end to end with a smart goal of technology, one that's realistic. How can I do this and do it in a very iterative process? We're too big to move quickly, so we gotta be smart and move slow.

Smith

Amy, legacy grants processes can be fragmented, burdensome and easily exploited. How can AI help mitigate program risk?

Fenstermacher

One of the challenges, particularly in the fraud and error space, is the fact that your adversaries are not static. So you're not solving the same simple problem, or complex problem even, every single time, you're solving a completely novel set of problems each time. The only protection we've had against that for decades has been clever people trying to outwit the clever people who were trying to take advantage of the system in the first place. So now, we have the opportunity to take that cleverness of our people and then to continue to codify and grow that, and let our systems effectively take on some of that burden with appropriate oversight. 

That's a critical component of this. But to take on some of that burden in order to escalate and identify potential flags for risk of fraud or what we call fraud waste or abuse. My thought on this is that AI really helps to mitigate program risk by ultimately differentiating between the high-risk scenarios and the lower-risk scenarios. In many cases, and for many programs, you see the granting agencies putting the same level of effort and the same monitoring activities against every single one of their grantees, even when some of them are known quantities that they've been working with for 20 years, have a pattern of strong execution, and some of them are brand new, and they're ultimately putting the same level of effort into both of them and following the same protocols. 

So AI gives us the capability to differentiate our monitoring protocols, so that for higher-risk situations, higher dollar value, lesser-known quantities, lesser-known recipients, that we can apply more effort and more human intervention and attention, while also allowing those who have proven themselves and have a pattern, and therefore are lower risk, to operate as normal without as much attention from the grantors. So, really AI is giving us the ability to focus our attention — our human attention — on the places that it's needed, rather than removing the need for that attention wholesale.

Smith

Okay. Just to move this on slightly, I understand, Amy, that you helped the Hawaii Department of Human Services rapidly apply US$80 million of federal funding to childcare providers during the pandemic. How did technology speed the process up there?

Fenstermacher

During the pandemic, the Hawaii Department of Human Services was working to get money out the door to childcare providers, but they had a particular challenge in their situation because, in Hawaii, the childcare infrastructure is pretty informal, and the legislation that was passed down by the government expected childcare providers to be highly regulated, certified in multiple ways, and then also to be institutions in and of themselves. 

But in Hawaii, it's mostly the lady at the end of the street who is taking care of all the kids on the street, and is certified and has her insurance and is legal, but is a lot less formal and has a lot less resources at her disposal. So, what we found was, first of all, we stood up our initial portal using that same EY Grants Accelerator and we set it up quickly and started running the program, running communications. 

We then nearly immediately discovered that there was very little uptake on the program, and nobody was interested in taking the funding, despite the fact that it was free and available. So, we discovered, based on that, that many of the childcare providers had a lot less resources, maybe didn't have desktop computers at all, maybe had two or three children clinging to each of their limbs, and therefore had a lot of trouble with sitting down and filling out an application. So, we updated the design of our portal to make it so that someone could fill out a grant application with one hand, thumbs on iPhone keyboard kind of experience, and could take photos of their documentation and upload it in the same movement to simplify the application process. 

We also found that, in many cases, the US government in Hawaii is not a comfortable funding source for many native Hawaiians. And they feel the remnants of the colonial history of the islands. So, we also needed to make a number of changes to the visuals and to the user interface to make it feel more locally controlled. Both of those changes could have taken months if we had used a traditional method of deployment, but because we were using our underlying low code platform, Microsoft Power Platform, it made it much faster, and we were able to turn around changes and adjust our approach within just a couple of days or a week or so. 

So, it was a very quick adjustment. And I'm proud to say that the Hawaii Department of Human Services has, as a result, been able to expend all of their funding and received additional over time. 

Smith

What a brilliant example. And, Andrea, just following on from that, what are the main challenges when using technology to standardize the grants process and enable scalable and coordinated collaboration?

Sampanis

We often see a challenge of cybersecurity sort of going against user experience — and user experience, a lot of what Amy just talked about, folks who don't have internet, who don't have computers, and those are the exact underserved communities we're trying to get to — or just the language we use. Sometimes we make them log in and, and because of security controls, they need to have a face scan, or a government-issued ID. Well, some of the people we're trying to reach don't have government-issued IDs. Immediately, we knocked them out of the process. 

So, it's something we really, I sort of not joke, but I use the three Ps when working and trying to solution: it's partner, pioneer and persevere. So, first, partnering to truly understand the real problems, understand why, just like Amy just said, why was the lady at the end of the street not applying for this almost free money that she could get? You know, it's something we had done wrong in the way we solutioned to try to get her to apply. Also, for me, in the federal government, I can't do this. We're a three-person team that's trying to create this. So we absolutely have to partner with folks and vendors like Ernst & Young, who are using innovative technology and also understanding their customers’ needs. 

The pioneering is the really difficult part of bringing that innovative technology that everyone wants to say, “Oh ChatGPT or AI,” but really the problem is, she needed something she could do where she could just take a picture and send it to you. And then the government on the back side uses AI to actually ingest that information and turn it into the high-tech systems we have. So, we have to identify what's the low-tech answer, because she applies for the grant, but, after that, she has to also report on the grant. And if they don't report on a frequent basis, then they get dinged and it's just this horrible cycle. 

So, we really have to think all the way through the grant cycle. And that brings me to the last one, which is persevere, because man, there are so many roadblocks, and I am sharing this because I'm hoping other grants officials in other countries are listening, because I'm sure you probably feel this way. You get down, there's so many roadblocks. It's like a layer of an onion, you just keep finding one every time. So you just have to stay dedicated and persevere and, again, realize it's going to move slow, but you can make incremental progress – as long as you partner and listen to the customer's needs, and you do pioneer with innovative technology. 

Smith

Finally, and I know you've both touched on this already to some extent, how can grant practitioners use technology to broaden access to funding and increase the impact of their grant programs? Andrea, let me put that to you first.

Sampanis

I feel like we may sound like a broken record, but it really is, it always goes back to those fundamentals, right, of making sure you are listening to your customers and you're streamlining that process. So you're making it easy for them to use, not putting up so many hurdles, like losing them from the get-go that they can't even log in. We use login.gov, it's a federal offering, security offer, to say, “Yes, I'm a human,” on the other side. But whenever we started this two years ago, it wouldn't work with anyone who was international. 

Some of the SMS texts wouldn't get to them, and then they would try to call up the helpdesk, which, by the way, didn't work at the hours of an international user. So, first, we had to fix that, make it 24/7, right, and then if they couldn't, if maybe they don't speak English, well, we had to make sure we created the translatable information. So, again, just thinking all the way around the user experience, streamlining that process. Ideally, we could use all the same one highly configurable low-code platform that EY brings — that's going to take us a long time to get there. 

So, in the meantime, using technology to connect it so it feels like a connected experience, and we share that data. I think that helps the grand tours, the US government employees, it's fewer systems for you to manage and fewer things for you to rekey and bringing it out of Excel and, again, partner, look for partners, don't be scared to talk to vendors and get ideas, and say, “I'm thinking this — what do you think?” Have that as a conversation. 

Smith

Amy, your thoughts on this?

Fenstermacher

One of the mantras you'll hear my team talking about internally is finding ways to access hard-to-reach people in hard-to-reach places. And I think some of that comes from mobile accessibility. Some of that comes from an improved user experience. We've done things like adding offline mode, but fundamentally, the way that you figure out how to get to new people or difficult to reach people is by having the data to help tell the story of who you are and are not reaching. 

So, we've seen over and over again that, when you lay out a map of where grant funding has gone, you'll see, effectively, fund recipient deserts, right, you'll see areas where, geographically, the dollars aren't going. And so that immediately helps you to start tailoring your approach to figure out why the dollars aren't going there and what you can do to get the dollars there. And then you can use those same data aggregations to evaluate whether your efforts are working. 

So, ultimately, it comes down to actually having the data to see the picture and tell the story, and then to evaluate your progress over time. So, it's not just monitoring the performance of the grant recipients, but also monitoring the performance of the grant programs themselves that becomes critical in actually achieving that goal. 

Smith

Okay. Well, that's it. We're out of time. Thanks very much for such an interesting conversation. Andrea, thank you for joining us.

Sampanis

Thank you so much. 

Smith

And, Amy, thanks very much to you. 

Fenstermacher

Thanks. Talk to you next time.

Smith

Do join us again soon, when we'll continue to look at transforming the grants process. And please subscribe to this series so you won't miss an episode. From me, Tim Smith, thanks for listening and goodbye.

Jingle

Granting the Future — back soon. 

Back to podcast