Use Predictive and GenAI for Fundraising & Prospect Research 

today to our webinar on utilizing predictive and Gen AI for fundraising and prospect research. I know we're getting into that sort of end of the year sprint when time becomes at a premium for everyone. So thank you for joining us, and we will get into it. So welcome, everybody. Just to give you a sense of who you're going to be talking with today, my name is Marissa Mabie, and I am a nonprofit customer solutions engineer here at Windfall. So I get the privilege of working with many of our nonprofit customers and have spent my entire career working either at or with nonprofits helping them leverage data and now data and AI. Olivia, do you wanna introduce yourself? I would love to. Thank you, Marissa. My name is Olivia. I'm on the new business side of of Windfall, meaning I help nonprofits move up into the right on the data maturity curve and help them evaluate data solutions. I think some of the biggest challenge we're having in 2026 on the nonprofit side of the business is understanding what solution you need, what data solution, what AI solution to fit your outcome. And my job is to really be the expert and guide you throughout that process. So excited to chat. Awesome. So, yeah, we are excited to jump in and talk about, again, the data maturity curve, like Olivia referenced. And, you know, no matter where you're at on that, how you can start to leverage high quality data and then supercharge that with predictive and generative AI to, increase your outcomes, save time, and, you know, become more efficient and effective, which is the name of the game here as I know we all have limited time between now and year end fundraising goals. So we're gonna talk also a little bit about, Windfall for those of you that aren't familiar with us and because all of the really, cool and innovative, you know, AI pieces we're going to be talking about are still based on a bedrock of having a very high fidelity, high quality dataset. That's actually where we got our start. Before we jump in and, you know, get into the content, we wanna ask those in the audience, you know, why did you join today's webinar? That'll help us really tailor our discussion as we go and make sure that we're talking to the, you know, areas that are most top of mind for everyone here. So I just launched this poll. Would love if you would just pick the answer that, or answers that best correspond to why you joined today. It's multiselect, so you can select all of them if, you know, you're really into... Here to learn about everything. Alright. Seeing some great participation. And I will end the poll and share the results with everyone so we can talk through this. So, you know, it looks like almost everyone who's here today really wants to figure out where AI fits into their fundraising, into your workflows, into what you're already doing. So, you know, Olivia, I think you're gonna have a lot to talk about when we get to that portion today. You know, another segment of folks saying, you know, prospect research currently in its current form just takes too long. Right? Maybe we can't keep up with all of the, you know, great leads and prospects that are coming in, or, you know, maybe we, you know, just wanna go faster in general. So we'll talk about that a little bit as well. As well as folks who are interested in overall just learning more about Windfall, our data products and services, and folks who wanna get ready for the great wealth transfer. Right? So how can we start to understand the velocity of how, you know, wealth and sort of charitable gift decision making is going to be changing, and how can AI help us get ahead of that? So good news. We're gonna be talking about all of that today, and I will get us on. So in terms of, you know, what we're targeting, what we're gonna talk about today, if you're a nonprofit professional focused on becoming more data driven and wanna leverage AI, wanna understand where it can fit in, like, welcome. You're with your people today talking with Olivia and myself. Today, we're gonna talk just briefly about Windfall's, you know, history and our product offerings, but then we're really gonna get into the generative and predictive AI side of the house, talk about use cases for data driven segmentation that can leverage AI, as well as how this intersects with the great transfer of wealth. We'll talk a little bit about Windfall's application, which is where a lot of these AI features can come to life. And if you do have any questions as we're going, we encourage you to leverage the q and a tab that'll be in the bottom of your Zoom tray. We're gonna try and answer questions live during the presentation as they come up, but we'll also leave time at the end to get to more or, you know, sort of continue the conversation. So we're probably not gonna have time to jump into, you know, very detailed information about our dataset and how we approach wealth. We do have other webinars that are far more tailored towards that, and I encourage you to sort of check out our events page to see what's upcoming if you really wanna dive in on that front. We also won't have time to get into our AI approach specific to your organization, your specific dataset. We're gonna be talking in somewhat of generality as here. But if you ever want to have a more tailored conversation, Olivia would be happy to help set that up with you. Similarly, we're not gonna get into, you know, pricing models and things like that, but, again, happy to have any of those conversations with you after the fact. So in terms of where we're going today, quick windfall overview and then jumping into of wealth, and then talking through tangible use cases with data and AI across predictive and also generative AI, and then talking about how this all comes together in the Windfall application for you. One more poll for you to help us again sort of gauge where everyone is at and how we can make best use of time today. So looking for info on what your organization is, you know, struggling with the most today. So let me just go and see if I can get that launched for us. And there we go. We should get the new question coming up for everyone as well as the answer options. This one is single choice, so we're making you pick, you know, what is best fit for you. Got some answers coming in. We will give folks a couple more seconds to lock them in, and then we will share the results with everyone. Alright. Lovely. I will end this poll and share the results. So we can see majority of folks, just slightly, are most interested in knowing which prospects to focus on. Right? That's what we're struggling with. And then we've got, I would say, a bit of a mixed bag across the other three, almost equal segments saying, hey. Our data is too messy as in... And incomplete. That's our issue. Or we don't have enough time for prospect research. Or, you know, we're really struggling with donor retention. So how do we start to keep donors giving year after year? Keep them engaged and keep them sort of building that sustaining behavior with us. So let's briefly talk about Some of the folks here, I think, are new and even said, you know, we'd like to learn a little bit more about Windfall as an organization. Before we do that, though, I wanna introduce the idea of the data maturity curve. So that's what I'm gonna build out here. So as we look at this, we're seeing how well you may have data available to support AI in action. Right? That's kind of the end game. That's what I know some folks have been, you know, tasked with. Hey. Integrate AI or, you know, better leverage it in your processes. But we actually wanna rewind a little bit and begin at the beginning and, you know, talk about, do we have data that we trust that is, you know, aggregated in a smart way that will point us in the right direction and actually make us more effective when we start to use AI on top of it. Right? We start to leverage AI. So we think about this in three different stages. So the first is the establishing phase. This is where most organizations start. It's where your organization still might be. It's where you're likely working out of a CRM, segmenting on transactions, maybe layering in some third party, maybe wealth data, other, you know, contact data, enrichment. You know, if, you know, some of your your challenges still get to basic segmentation or reporting, then, you know, you may be here. When we think about being in the driving phase, that's where we are starting to get more value out of this same data. We're able to do smarter segmentation. We have better reporting. We may have tailored, dashboards and views that really help us dive into what's top of mind for, you know, frontline fundraisers or for folks on the prospect research and qualification side of the house. And, you know, we may also be leveraging predictive models at this phase. So, you know, that's where we may be looking for guidance around how we can better leverage our data and AI in our workflows. Right? We may be at that stage where we're really starting to get some lift out of all of this data and out of, you know, having workflows and predictive AI suggest things for us, surface up top prospects for us, and kind of do some of that heavy lifting. The last stage here is where we get into the orchestration stage, and this is where everything comes together. And this is where our data can start to trigger actions automatically. This is where we start to get into things like generative AI, where we can start to generate content or suggest, you know, the best times to, outreach to specific prospects based on past giving patterns, start to suggest language for outreach. Right? AI can generate content. Systems can alert us when something changes when we may wanna take action, and we can start to create prescriptive workflows within our CRM or other tools to help make all of that happen and help us connect the dots. And so if what your, you know, challenge is is really around getting data embedded in your decisioning or your existing AI workflows, you're probably in this stage of the maturity curve. The key point here is no matter where you're at, this isn't about being at the right place, being at the best place, being the most advanced, having good or bad answers. It's just about identifying where data can start helping you do more of the work in your organization where you can really start building this engine that then we can start to use generative and predictive AI to, you know, build on. How do we string this together and how to create... We create a strong foundation? We spend so much time talking about this and even talking about data because understanding your constituents is harder than ever. And we've heard this since our inception in 2016, and it's still true today. It's a little bit different today, but it's still true. So, like, there's never been more data out there that we could potentially, you know, get or leverage on our constituents. But we still face a very high barrier of having to go through and figure out, well, what's correct? What's good data? What can we string together? What's significant? And that takes a lot of time and money and, you know, person power to sort of comb through. So that's really where we got our start coming in to help combat these data challenges. So our mission at Windfall is to change the way that organizations understand and engage with their donors by helping them leverage accurate up to date information. And we do that by being hyper focused on data quality by providing back a precise household net worth down to the $100,000 figure for every record we can deterministically match from from our customer's dataset to our own proprietary underlying dataset. And that's really key for us because we own our own dataset. So we control the quality. We control the enhancement road map, and we also control the update cadence. So we actually update our whole over a 110,000,000 household dataset on a weekly basis. Because if you think about how quickly wealth creation and destruction events or signals are surfacing in, you know, the public realm and how often the data is changing, there really is a lot that you can capture on a weekly basis. So that's how frequently we're rebuilding this and then providing our customers with new data, updated data points, and some of these great time based triggers that we're gonna talk about as we get a little further into AI workflows. Today, we work with more than 1,500 organizations, many nonprofits across the country, across the, you organization type, across organization, size. And, you know, just like it doesn't really matter where you are on the data maturity curve, there's a way you can start to better leverage data and get more from it. You know, if you're a small organization, no matter your CRM that you have stood up, like, there are ways that you can start to leverage predictive and generative AI to work more efficiently and effectively. So Windfall really empowers our nonprofit customers with accurate data. So I mentioned precise net worth. That was the first and only data point we used to provide back when we got our start, But that's very powerful as you can imagine because that really helps you start to stack rank your database and start to understand, you know, who are some of our highest potential donors based on net worth, who should be in a portfolio, who should we make sure that we are aware, you know, is in here or has recently engaged with our organization? Next, we think about segmenting or finding hidden gems. And this is, again, where we use net worth. But then we start to pair that with how folks have already engaged with your organization. Right? You've already, usually, if they're in your CRM, sort of got them to maybe donate once, got them to sign up to show that they're interested in your cause, that they want you to contact them. So how can we start to compare how they've engaged, their gift size, things like that, with their net worth, and understand, like, do we have them in sort of the right solicitation pathway? Are they, you know, unassigned sort of in our annual fund pathway where, really, they look like a major gift prospect, and we may want to speak to them one on one and see if we can get them from giving, you know, small gifts to sponsoring programs or to, you know, really getting more enmeshed in the organization. The last piece here is around engagement. So, you know, once we start to understand net worth and past behavior, there's still a lot of ways that we may want to splice our data and start to segment folks for outreach that's going to resonate. And that's where the now 50 triggers of additional information that Windfall provides back on matched households can really help you take that personalization and outreach to the next level. So we call these Windfall triggers. We'll talk about them on a couple slides from now, but they help us really understand, you know, what a household may care about, how they behave, temporal events that they may have experienced that may make it a good or maybe not so good time to reach out. So when we pair all of this together, we start to have the underlying ingredients for creating, you know, very smart data driven, and eventually AI driven workflows that can help us segment folks or pick up on changes or events that may lead us to want to reach out. So Windfall's platform and sort of our product offerings are built for data driven development, but everything goes back to our core wealth screening. So that's where we're integrating either, you know, with your CRM or utilizing CSV files to match the constituents that are in your, database. And then when we do make those deterministic or very strong multifactor matches, then providing back the enriched data. So a precise net worth as well as up to the other 50 triggers of information to help you really understand who these folks are and how to best reach out and resonate with them. We also leverage predictive AI. So this is where propensity or propensity to give modeling comes in. So this is based on our existing data infrastructure as well as our deep bench of data science team at Windfall. This is where we work with customers in a more bespoke fashion to create propensity scores based on your first party data as well as Windfall's data. So we work together to scope out, like, exactly the kind of donors that we want to help identify more of or the exact actions we want to predict that people will take, and then we build models that help you understand against your whole database who's gonna be most likely to do that. We also offer data link, which is a product that really helps organizations that may have disparate data across different databases, different CRM structures, health care organizations, ticketing and membership organizations, lots of uses here to really wanna understand who someone is across organizational silos or systems and understand how they're behaving and engaging with your organization in aggregate. Something we always like to call out is that our pricing model is not based on a number of records or credits. We provide unlimited syncs for all products. And providing unlimited syncs is actually a little bit new. So when we came on the scene to... And even, you know, still typically, you know, other vendors may have you only select a subset of records to screen. Right? Only recent donors or only folks that are, you know, giving above a certain level or already assigned to portfolios. You know? However, right, you're working within a constrained system, not screening everyone. And no matter how you slice and dice it, if you're not screening everyone, you're going to miss potential affluent constituents that, you know, could potentially be giving at a higher level to your organization, could be a great fit for major gifts. And so that's why, you know, we said, let's let's not mess around with the credit system. Let's just let you screen everyone and then rescreen everyone. Right? Because things can change. Wealth is very dynamic, and we wanna make sure that our customers always have the most updated net worth and other triggers on their constituents that they can, you know, leverage to make outreach and messaging decisions. In terms of the exact data points that we provide back, so the wealth screening attributes are showcased on this slide here. You can see there's about, 30, three, 34 of them here, so provided back for every constituent that deterministically matches from your dataset to windfalls, and all of these coming in at the household level. Our newest attributes I wanna call out here are the donor advised fund attributes. So is the household affiliated with the donor advised fund? That's obviously very timely as we're thinking about, you know, DAF Day a few weeks ago and also planning for end of the year giving and sort of tax advantage giving strategies. And also interest in crypto. These two triggers are part of our premier package, which is what is showcased on this slide as well here. So it has those two triggers as well as career intelligence data. So this data flows in at the individual level and really helps us understand not just a picture of someone's, you know, affluence, but a picture of, you know, what they've spent their life working on, what they're likely interested in, what their influence level may be, you know, professionally. So this can be really powerful when paired with the... Both data and also when paired with your internal engagement data. Alright. So now we've talked about the, you know, availability of data, some things that Windfall provides back as well as, you know, what else is out there. So we're very interested now when it comes to segmenting your own data. How... You know, are you currently doing that, and would you say it is easy or, you know, maybe more difficult than you would like it to be? So I believe we have launched this poll. And thank you for folks that are already starting to respond. Alright. I'll give about five more seconds here, and we will end it. This is another one we forced you into a single choice for. So folks picked the answers the best corresponded with where they're at, and we see a almost even mix across folks saying, hey. It's easy to segment our data. Or, yeah, we do it, but it's mostly manual. Gotta love, you know, Excel or SQL or whatever you're you're using on that front. Some folks saying, we segment, but mostly just by past gifts or purchases. So maybe things that are very, you know, easy to get in and report out through our CRM. And then we do have some folks who said, hey. We're not, you know, segmenting yet, but, you know, here to here to learn. So thank you for sharing that, everybody. So now we're gonna talk about the transfer of wealth and some of the data and AI opportunities that it really opens up that we want to highlight. Right? So, let me get us into our next slide. So what we're looking at here what we're looking at here is US household wealth, and how it's been sort of moving across generations, across the last twenty, twenty five ish years. Right? And we can see, you know, up to about twenty twenty four, 2025 that the silent generations and especially baby boomers are still holding the lion's share of wealth in The US. Right? Trillions and trillions of dollars with millennials and and Gen Zs kind of kind of down here with a smaller portion. When we look at how that is actually going to change when we fast forward into the next twenty years, you know, we see why people are calling this the really, you know, great wealth transfer, and we're actually already in the midst of it. So over the next years, we're going to see over a $100,000,000,000,000 be transferred from, you know, baby boomers primarily down to Gen X millennials and Gen Z and younger generations. And we already know that, you know, out of the $124,000,000,000,000, 18,000,000,000,000 of it is already earmarked to go to charities. Right? Maybe people signing the giving pay pledge, putting money into donor advised funds or sort of other sorts of trusts and things. But there's still a a lot of money that is going to change hands and is going to, you know, change hands largely within, you know, familial units and things like that. So when we're thinking about, well, who should we be talking to in our, you know, CRM? Who may represent the folks that may be the future decision makers for a lot of this wealth. It may not be the folks that are currently coming to the top as your most affluent donors or your most engaged donors. We also like to pair that with the fact that overall US consumer wealth, when you sort of even net out the generational, lens, is also, you know, changing a lot. And in this case, we're looking only at, like, the past five years. And we're looking at the percentage of total US wealth that is held by high net worth households or households that have that precise net worth of a million dollars or more. And we've seen this continually since, you know, COVID tick up and tick up and up until we get to, you know, today, in 2026, where we're looking at almost 81% of wealth is held by high net worth households. So it's starting to be, you know, very concentrated in these high net worth households. And it's also increasing overall. Right? And high net worth high net worth households themselves, if you look at the number of them, are also increasing in The US. So if we're, you know, looking at our database and saying, hey. We want to focus on folks that are worth a million dollars or more. Those are high net worth folks. We may have more than we know what to do with. Right? We can't prioritize them all. We can't put them all in major gift portfolios. And that's where, as we start to talk about leveraging generative and predictive AI, we can really start to make smarter and more prescriptive descript... Decisions. So last slide I'm going to talk to you here before I pass it to Olivia. This is where we start to ask ourselves that question of who should we focus on in our database. Right? We have prospects. We have constituents. We have lapsed donors. We have recent donors. We've got a lot. We may have millions of records. How do we prioritize them appropriately? So, you know, as prospect research folks or, you know, development folks, we're trying to find each single person. Right? We're trying to, like, flip over all of these rocks and say, who's gonna be most likely to make a gift to us or make a major gift to us? And, you know, how can we start to stack rank everyone, flip over all of those rocks, and, like, really understand who we wanna focus on first, who we wanna put into a portfolio? And that's really hard to do, you know, one to one through everyone in your database, even through everyone who's affluent in your database. So this is really where data and AI can start to help us, start to help us build those, you know, predictive... Protect... Propensity models and understand who's going to be most likely to respond or most likely to make a major gift and really help us, instead of driving ourselves crazy by trying to, you know, deep dive into every single person in the CRM, understand where our time is best spent and where we should be focusing. I'm gonna pass I'm gonna pass this to Olivia. Everything looks good, Olivia. But if you're talking, you're on mute. Awesome. Thank you, Marissa. Alright. So before we... Yes. We have setting the stage, the analytic maturity curve progression. But before I explain this to everyone here, there's two stats that Marissa just mentioned that I want to reiterate because I think they're very important for every prospect researcher and development professional. The first is that 80... Over 80% of The US wealth today has a net worth over $1,000,000, and that is only growing larger. So all of a sudden, right, if you're a prospect researcher, you have this massive pool, right, of of wealth that you are now leveraging. But what were happening, what you're seeing with the great transfer of wealth is all of your historical major gifts, right, that year over year you've been able to rely on, when they... When that generation dies and, you know, passes on, their philanthropic behavior is not going to pass on as well to the next generation. Right? That, the Gen X, the millennials, they're gonna have a majorly different view of how they think about philanthropy and who they wanna give to. This is why it is so important to start leveraging AI to uncover what is that next generation and what do they care about so that when, right, that wealth gets transferred, we are already engaged. Right? We already have that relationship to, right, kind of, to qualify for their next major gift. So you're not having that major delta on your giving. So before we dive into AI, let's set the stage for this maturity curve progression. Here you're gonna see on the bottom left, descriptive analytics. And as we move up into the right and into this analytic maturity curve, this is how you're gonna really start optimizing, surfacing those prospects. So descriptive analytics, you know, has been around since ancient Greek and ancient Egyptian times with understanding what happened. Right? This is kind of taking that census, taking a population view, under... Like, looking at, you know, how many gifts did we receive this month. As we move up, we wanna understand why did this happen. Right? Why did we receive a 100 direct mail, gifts just specifically in this ZIP code? Right? We wanna understand why this action took place. The next kind of cohort here is gonna be predictive analytics. So this is where I actually think AI holds the most power for nonprofits. This is understanding what cohort is going to be your next major giving cohort. Right? What's going to happen next? What's gonna happen in five years? Who's gonna be my next plan giving prospect? Who's gonna be my next prospect who's gonna give $5,000,000 to my hospital or, university? This is really where we're gonna focus the meat and potatoes in that predictive analytics cohort. But as we move on, right, kind of prescriptive analytics, this was the dawn of generative AI. This was the dawn of chat GBT, large language learning models, when we can really take these large swaths of data and say, alright. We kind of have this. What predictive analytics has done is they're giving me... You know, here's the top 100 prospects who are gonna be most likely to give a million dollars and above. What prescriptive analytics can do is they can say, you're gonna take these top 50, and you're gonna call them, and they wanna respond for with a, you know, a female voice in a phone call and this is the exact amount you're gonna ask. Or you're gonna take this a 100 and they're gonna respond best to a direct mail in this specific zip code having this be the minimum bar. Right? It's going to really drive home how to engage very specific, prospects for stewardship. So these two, right, is what we're gonna focus. Now one important call out here. All of this AI and propensity modeling and, you know, generative AI creation, all of it depends on having a high fidelity, highly accurate dataset that you're working with in the first place. I think this term has been overused a lot, but it but it remains true. Bad data in, bad data out. Right? This is kind of the caution as we go through this. You know, when we're kinda taking this back to that maturity curve that Marissa went over, we wanna make sure we're working with the correct contact information, the correct first party information as we do this appendant enrichment, of your database and donors. So what exactly is AI? Well, I I always like this slide because it reminds me AI has been around longer than we think. AI didn't start with large language learning models. AI started about over twenty years ago with the dawn of machine learning. Right? So it can mean a lot of different things depending on the context. But, again, the few... These are the really important high level bullet points of what AI means today. First and foremost, machine learning, and that is predictive AI is a subset of machine learning. We'll dive into the specific models, but, right, this is supervised learning models, regression models, tree based models. All of this had been around for over twenty years, and it allows us to make that predictive what will happen next type of analytical view. The second here is you have, right, the natural natural language processing. These are called NLPs. These are really great for translation or classification modeling, speech voice recognition. Right? That's also part of the AI bucket as well as what has spawned over the past, you know, five years, which is large language models in GenAI. And those are gonna be your chat GBTs, clods, etcetera. Alright. How do they all fit together? This is basically everything I just said, but bucketed in a really nice infographic. So you see here, right, AI encompasses all of these three things. It's a program that can sense, reason, act, and adapt. Machine learning is a subset of AI. These are algorithms whose performance improves as they are exposed to more data over time. We're gonna get deeper into this, but that's a really important note right there. Right? Machine learnings are models that improve over time. Deep learning, this is a subset of machine learning, which you have multilayered neutral networks learning from vast amounts of data, And then you have your smallest subset and your newest subset, which which is GenAI. And this is for advanced applications of deep learning that create new content by learning from data. This is really where that content creation piece comes in, and we'll dive into how Winfall thinks about this in a bit. Feel free to screenshot this. You know, this is really helpful for me to understand, right, just how broad of a subject AI is. Alright. So where does predictive AI fit into understanding prospects and kind of how Windfall thinks about this? So I wanna... Your eyes to first start here. Right? We've... Marissa talked a lot about our net worth. And and wealth is important. Right? Wealth is is a critical piece of someone's portfolio to understand their capacity, but it doesn't paint the full peck... Picture. In fact, it's almost, right, kind of a third of the picture here. Because John Smith on the left could have a much higher net worth, but when we look at someone's propensity to give score, right, their ability to actually... Their most likelihood to give action, he has a much lower propensity to give than the constituent on the right. This is because we're leveraging supervised learning models to actually understand propensity versus just wealth. This is the benefit of what you can achieve when you add on and layer on propensity modeling to your organization. It really takes that large swath of, okay, someone might be wealthy, but so what? I need to know if they're actually likely to give to my organization. Right? This isn't just giving in general. It's giving at my organization to a specific cause. So there are many use cases for this type of AI modeling. We have organizations who... They just need to understand who is most likely to to participate in a annual fund. They don't care about the dollar amount. This is really critical again, when you're trying to target that next generation. We just care about engagement. We don't need them to be giving, you know, more than a dollar or 500. That's not the goal here. The goal is to get them... Is optimize engagement. This is what we can build AI models to help predict. The second is we actually do care about leadership gifts. Right? We wanna actually generate a dollar amount and look for people who can give in amounts greater than a thousand. We really wanna focus on this a thousand to 5,000 category. Major gifts. We wanna move up into the right. We need to focus on constituents who can give in the next three to five years $50,000 or more. Right? We're gonna actually build a time frame as well as a gift threshold. Grateful patients, this is specifically for health care, right, when we wanna understand, what patients are most likely to give after, you know, a visit, at a specific dollar amount. Planned giving. Right? This is gonna be really critical when you're looking at that silent generation, that baby boober generation. Let's look at that massive pool. Right? They represent... They're holding all of the wealth now. We wanna build a propensity to give model to capitalize on looking at who is most likely to give an estate gift or a bequest based off of, right, all of this first party data and wealth data. We build these models and many, many more. We build direct mail optimization models. Again, this is the meat and potatoes of where AI lives today, and where you're gonna see the highest return of investing in AI is through building these predictive models with having these bespoke outcomes specific to your organization. So we're gonna... I'm gonna say this word a lot, bespoke. Right? This is really the power of working with Windfall is because these models are built specifically for your organization. You could have two different... Two of the same constituents living in, you know, your competitor's nonprofit. They're gonna be receiving drastic different scores, because we're looking at your unique first party data. So, really, when we talk about machine learning and predictive AI, the first step is coming up with an idea. That's your idea. Right? This is what you need as a business. You need major gift, pipeline optimization. You might need need direct mail optimization. Whatever it is, it's completely customized to your goal. Wonderful. You become up with the idea, and then we build the model. How we build a model when we talk about predictive AI is we look at all of your first party data from, past giving information, engagement data, event attendance, any type of marketing data you have on your constituents. And we do an exploratory data analysis where we actually combined all of that first party data with Windfall's underlining data asset of over a 100,000,000 US households. So what that looks like is it's an extensive deep dive, right, into understanding what is the ideal makeup of the prospects who are going to achieve your desired outcome. We're then gonna choose a specific algorithm, right, of what we think suits best for this specific model, and then we're going to run it. Right? We're gonna score every single person in your database or CRM on a scale of zero to 99 of how likely that constituent is to achieve your desired outcome. But we don't stop there. Right? This isn't a one and done. As you can see here, this is a loop. And if you remember what I mentioned before, right, these models get smarter over time, these machine learning models. So what we're gonna do is this model is going to change in real time as new data comes in. So I'll give you an example. It's one of my favorite examples. We worked with a customer last week where one of their constituents scored an extremely high. It was an 80 out of 99 for a major gift prospect. But all of a sudden, right, we registered a divorce. That this household in real time, we registered a divorce, you know, three days later, and that score dropped from an 88 to a 40. Right? Because that's a major wealth destruction event. They didn't need to rescreen. They didn't need to do anything. Right? It's an automatic update because these models are real time, and they will get smarter as we take in donation data or no donation data. Right? They're going to adjust. So that's the beauty of leveraging these predictive models is you're getting scores for all of your constituents, and then they update on a regular cadence. So it's gonna stay current, in real time. So this is also, I think, one of a a really powerful slide because a lot of the time when you get these scores from, you know, other vendors, if you've done a modeling like this in the past, it can be a black box. Right? And a lot of the time you might not know what has actually gone in to creating these scores and having, you know, what data has an impact on the constituent doing your desired outcome. So with Windfall, how we think about this is through feature importance. So feature importance is a part of the machine learning data, model creation process where we wanna better understand the model's logic and verify that this model is making sense. So how we do this, right, is we... The model ranks all these features. So you can see here, this was a... You know, this is kind of a a dummy data here, but donations over the past five years was the most important feature in someone doing a desired outcome. The second most important feature is event... Oh, sorry. Oh, oh, is event attendance. Why this is important is because now all of a sudden, this data becomes actionable. You just built a major gift model. And what this major gift model is telling you is that it's really important when someone attends an event, it makes it even more likely for them to make this gift. See, I would then take this and go back to my marketing team, go back to my VP of marketing, go back to my CEO and say, we need to double down on events this year. This is a major indicative data point that is showing... You know, this is... It's a it's a huge feature that is driving the outcome that we need to achieve. And so these are things that you can start doing, right, when you're taking that deep learning approach to data and having these results bubbled up in real time. So this is... That's how one fault thinks about machine learning and modeling. Right? And this is, I think, a nice visualization of what this kind of looks like in practice. So we automate your data and AI efforts by setting up these flows in real time without you having to, you know, manually export CSVs or have it be, more manual than you would like. And so what this looks like is we're gonna be automatically connecting right to your constituent data, event data, volunteer data, donation data, engagement data, etcetera. And what we're able to do is not only enrich all of that data with wealth screening and the career intelligence that Marissa went over, but we can also help with that propensity AI and those predictive models. Right? And what that ultimately leads to is creating donor segmentation, portfolio allocation, win back campaigns because you have all of that wealth, career, and propensity AI models in one place, right within your, source of truth today. So we have built out an application specifically for nonprofits that allows you to do all of the segmentation and surfacing of hidden wealth in real time. Think of it as a software that lives on top of your CRM and combines all of your first party data with, Windfall's underlining data asset. Now, yes, you can create really powerful hidden gem segments, win back campaign segments, direct mail optimization segments within this application. But I actually think the most interesting part of this application has provided a lot of our nonprofits with some of the highest leverage is dossier creation. So this is, again, what Windfall thinks about when we talk about generative AI. It's how can we create specific recommendations and summaries of constituents, based off of these large swaths of data and be really, really prescriptive. Right? That's gonna be that that... On that analytic maturity curve, that top right banner. So this is something that we offer, that we've released, my god, I wanna say now over to... Maybe two years, over a year and a half ago. So really with the dawn, right, of generative AI, we were part of the wave where we said, alright. We're gonna help our nonprofits. We get very granular understanding of the best way to engage with Cynthia. Right? We're gonna take personalized talking points generated from a constituent's wealth, philanthropic history, and behavioral signals so that when we generate this dossier in, you know, less than thirty seconds, we can send this to our board member or we can send this to our major gifts officer who's about to meet with, Cynthia or our board member who's gonna be sitting next to Cynthia, at a VIP dinner table, right, and have this beautifully, crafted dossier prescribe, the best form of outreach for an individual constituent. And so what you... What you're kind of seeing here is this one two punch. Right? You have predictive AI and those propensity to give models to really hone in on, you know, taking a list of a thousand prospects and generating, like, the top 10. And then you use the dossiers to layer on, okay, we have these 10 prospects who have the highest propensity to get my major gift level. We're now going to run dossiers on these 10 to get really prescriptive on how to personalize outreach for them. And that, in a nutshell, is how windfall thinks about AI.