View Recording: AI-Powered Next Best Actions in Financial Services

We are really excited to talk about AI in next best action systems. Today I am Nathan’s noski am concurrencies, chief technology officer, and we’re going to be talking a little bit about how this has been applied in the financial services industry. I am glad to see everyone here. What we’re going to start with today is I just want to get a little perspective of where you are in applying next best action within your business. So I’m going to put a poll out there. You can’t hear me. Hopefully other people can. Amia might come through, OK. Amy Cousland 1:03 Yeah, you’re coming through for me, OK? Nathan Lasnoski 1:05 OK. Thank you. Great. OK. So we’re gonna start this off by having a little bit of a poll of where you are with your next best action implementation. Say that 10 times fast, so I’m going to launch that pole and the options that you can select from are. I’m not sure what it is. Why am I here? I’ve never heard of next best action before the 2nd is I’m exploring it and haven’t put it into production. Haven’t really started to really get any real wins out of it. The 4th would be. I’m already have some part of it in production, but I’m not using AI, it’s more statically driven and then the last is we have an AI solution in production, which if you’re in that point I would love to talk with you more. So please fill out that poll, and I’d love to. Let’s see where you guys are at already. I’ll give you. I’ll give you about 10 seconds. 10 more seconds to fill the sound. OK. Thank you. Alright, so looks like about 20% of you are not sure what it is, so we’ll make sure we define that for you. Thanks for answering that way. Umm, we we have about 40% of you that are exploring it and another 40% that have started to put it into all that’s changing already 29% that have already have some part of it in production without AI and 14% already have an AI solution in production. So I’m really looking forward to learning more about that solution as well. So awesome, this is great. So this is going to be a session where you can feel free to drive any question you want into the chat. No, provided it’s not about whether you can hear me or not. Just kidding. Please make sure that you asked those questions. We’re gonna answer them as we go. We have a variety of different concurrency members on this call. In addition to myself, so we can answer those real time and I’d love to have them come up as we hit different parts of the presentation. So feel free to let them fly. So let’s start by just defining what next best action scenario is really are I think, an important way to think about this is thinking about the capabilities of AI and the context of what we can do now and what we are going to have the opportunity to do. And next best action system is where you have the opportunity to be able to define and especially in the financial services space, to define what the preferred action is based upon a set of criteria. So let’s imagine that you have a financial services advisor and you know that that advisor has a customer with a portfolio and that portfolio indicates that they’re not going to have enough money by the time that they retire and the results of that is that they’ll run out of money and they’ll be, you know, 75 years old and have already exhausted their available funds. What would you prefer for that financial advisor to do at this given point with that customer and maybe even what is the customers next best action at that point? Is it to increase their savings? Is it to increase the amount they put away into their 401K or to another savings vehicle? And what you might experience within your organization is that there’s a diversity among the choices that many of your financial advisors or your customers would take upon that point. But you know that there’s certain paths that are intentionally the best for that customer to move down. So what next best action is about is optimizing both your teams journey at helping your customers to take appropriate actions. It may be unearthing journeys that they’re unaware of. Think about things like multi relational families. You have different tiers of relationships you might have. You might have situations occurring that the financial advisor isn’t even aware of, or you might have inexperienced advisors that don’t know how to guide in a particular situation. So your goal is to help them. So how does AI help us in that scenario? Well, the ways it would help us is by predicting that something’s going to happen, like I’m going to run out of retirement funds and then prescribing what to do about that. And those are two situations that we’ve been using AI for a while for. These are not necessarily new to financial services, but they’re opportunities that continue to become more efficient and more capable every day. So the first step on that journey is we’re predicting something based upon a data set. So I know something about a person’s financial position. I’m predicting what’s going to happen based upon their current amount of spend or their predicted spend with that data set. So again, the situation is the person is not doesn’t have enough for retirement savings, for their level of income, their level of monthly spend and what’s going to happen after they pull the trigger on retirement. What you meant might then do is you might go into a pattern where you then prescribe what to do about that, and we just talked about the situation like prescription. So you might say increase your 401K or you might say, reduce your spending might be another prescription and you need to enable that financial advisor to apply that advisement. Once they know about the situation, you’ll notice that these two activities are different than what we might do with normal analytics. A lot of analytics situations we give data to a person, but they don’t know what to do with it or they don’t even necessarily know what it’s predicting. So our goal with AI is to take a step beyond analytics and help our advisors help our customers help our sales motions to be predicting what the right customers are to have conversations with and to prescribe what to do about it. But then what? Where it might go from here is to story tell on those larger data sets. So storytelling in the context of AI might mean based upon this situation in these prescribed Actions, what would happen if I did X? And those are scenarios that you can help to model for your advisors or help to model for your customers. That might even do it themselves in the context of your online assets. So storytelling is that next step around AI that enables us to help them to make strategic decisions, not just tactical decisions. So then based upon those things, that sounds like a pretty sort of taught understanding of next best action. So what are the rest of this stuff? So what we’ve started to emerge from in these first activities, these have been around for a while. What we started to see is that general activities have started to become a thing in the context of AI, and this is really where GPT models have made some of this conversation substantially more capable. So general directed activities, meaning activities that are more general in nature, but they’re still directed by a person. So it might be something like. Here’s a report about my customer, my person I’m advising return back to me. Things like the net income and what the evaluation score for their retirement plan is and what their investment mix is and whether it’s appropriate for the period of time. And you can start to inform that to think through that process a little bit closer to the way that a human might think through that process. So it’s like deal with general directed activities is something that has become a thing with the advent of large language models, which lets us start to move into directed autonomous activities. So an example of this in the next best action might be all right. I already know that I might be able to look at it particular financial plan and provide some perspective on what that person should do. Well, maybe now that I know that the boundaries of what that is, I can say go look at my thousand different financial plans and come back with the 15 customers that I need to prioritize this week to have a conversation with because of what it’s predicting about their financial ecosystem and that’s causing the AI agent to do something that perhaps a person might have spent a fair amount of time on in preparing for the financial advisor or maybe they wouldn’t have done it all, which is oftentimes the case in order for them to normalize the activities. And at which point you start to move into this idea of creative directive activities, which is about how should I change My Portfolio. And you already noticed that some of the even GPT models can start to make intelligent decisions about changes to portfolios or changes to mixes of investments based upon the intuitive nature of that. But you might even get to a point where some of that falls into the creative category. Where where a human has to has traditionally had to think about that as part of their their journey. You’ll notice that AI tools are starting to help the creative journey for an individual as well as they’re thinking about the context of the portfolio and how they can help an individual. So that’s essentially where we land now where we’re going is this idea of general or creative autonomous activities being part of the AI skill set as well. Now, that’s not where many of these scenarios exist yet, but it’s definitely where we’re going to get to where I portfolio optimization that was once not possible through robo advisors and bot driven experiences are now becoming a capability of what we have the ability to deliver through our AI platforms as part of the journey that we’re going through now. You may start down here just to get through the blocking and tackling and helping to provide value back to your customers, but know that there’s a journey that you can move through as you’re starting to think about this. So as you have questions about this particular breakdown, put them in the chat and I’ll hit him as we as we go along here or I’ll hit him at the end if if there’s a situation where you have some more questions. OK, so let’s let’s take a step back for a second. All that’s interesting. Let’s frame up the journey. Just a little bit before we go into into the specifics of next best action. So as you’re starting to leverage AI in this ecosystem, I always think that this is an important thing to think about is that many companies have already started right here. They’re already starting in a value unit scenario. They’re already thinking about how they can apply this capability to their customers or their employees value streams. It’s really important as you are engaging in your AI cycle to bring this back to the executive level, because this is truly a business changing capability within your organization. Even if you’ve done next best action before and you’ve done it in a way which is very sort of rules driven, you can get to a point where it expands upon what a rule is driven engine can do. And the reason why the executive alignment is so important is this is going to start to change in a really dramatic way. The jobs of all your team members and how they engage with their customers and how they optimize the outcomes for their customers. So it’s really important that you back up the train, get to the executive alignment of how you were going to leverage this across your company and then you bring that forward into where these capabilities can be driven into pitoc and pilots and production. And ultimately what this looks like is a scaled pattern across the organization, but reminder always to pull that back to the executive, the executive organization. What we’ve seen is that companies that try to pursue some of these too fast without having that executive alignment, they’ve run into challenges at POC purgatory and they’re not able to get it pushed through because the business is in the line. The business isn’t part of the adoption cycle. So as we start to pass this forward, one other sort of framing thing, I want you to think about is that in your AI solution space, there’s a variety of tools that are gonna fit into this picture. And next best action really isn’t an sort of immune to this. Many of the things we’re doing in the next best action will fit into simple commodity capabilities that are available through tools we already have. There’s platforms that let us take a look at our financial portfolios today that are going to start to develop AI tools, but many of the competitive advantages that companies are working on are ones where there’s direct investment necessary for you to gain ground against other companies in the same space. And we call those mission driven ones where you’re investing above and beyond what the commodity market is providing to you to enable outcomes that are specific to your organizations. Investment portfolio choices, your philosophy, the way you think about this picture, and I think that’s where you’re going to see delineation is certain organizations have certain investment philosophies or even customers have certain investment philosophies where next best action is going to be specific to a customer, not just specific to an organization. So a few things that you might be thinking about as we’re talking about next specs to Action and AI specifically, are these blockers and I wanna hit these. You know, I hesitated. I’m like, do I wanna cover these in this section? This is sort of a generic topic like I know everybody kind of wants to learn about next best action specifically, but every time I talk about one of these scenarios with a financial company, I was just with an executive team yesterday talking about it. This is the thing that comes up. First, they all wanted to understand how they clear some of these blockers, so I want to make sure that this is something you can take home and have as an asset for you. The biggest concern everybody has is around the context of data privacy and part of that is the effective fear, uncertainty and doubt. That’s very true around some AI platforms like what’s your some of your financial advisors may even be doing is using chat, GPT or putting things into the Google search interface that suddenly become part of the trained Internet. And that’s a really problematic situation. So you’ll notice that many tools that can do textual analysis or visual analysis to help optimize a portfolio or a choice or something based upon a document can be used very quickly to like even without your ascent, your acceptance to be able to do next best action. So it’s really important that you be very intentional about what tools you’re Employees are using to do this, A because you wanna make sure that there are low hallucinations or low end that the choice is aligned to your investment philosophy and B you want to have data privacy so you can stand up to the regulations that you’re held accountable to. So it’s very important that anything you’re building is using private instances of AI models, so that is something to hold your vendors accountable to. But B, also something to hold yourself accountable to as you’re building these solutions. So data privacy concerns really come into place. It’s actually not that difficult to align to because most of the solutions like Azure or Microsoft or an FIS or something like that, they’re already building around some of these controls, but because it’s something that’s evolving every day and especially for some of you that might do work in the European Union and other places, some of this is evolving in a way that we have to be very conscious of. So data privacy concerns that comes into play, and it’s necessary to use private instances and have walls that mitigate nonprivate instances from being used. The second use case is in this scenario of data readiness. This is a little bit of a boogeyman, right? Like sometimes we’re like my my data not being ready. It can be like a three year journey and then you can go to the ball like this is more about a specific use case with specific data sets understanding in the context of building up financial planning or other tools that it we’re directing our customers on. What am I causing them to? What data do I need to be successful at that historically, purposely? What data will have successful outcomes? What data will not have successful outcomes and then especially like what data should be used to accomplish an outcome and not get a little too bullish on how that data might be used so data readiness is more about functional data readiness on a specific use case specific value stream, then it is about the entire the entirety of the entire like set of your organization’s data in of itself, and that’s that sounds sort of obvious to say, but oftentimes it’s not obvious to people adopting AI of the third scenario and we’ll talk about this a little bit. More later is this idea on human displacement? Umm, I really recommend that Executive teams turn that comment around from exec human displacement to human Enablement. This idea that AI is a force multiplier that every individual within your organization should have the power to be more as a result of using AI. And that’s very true for financial advisors, especially ones that may be generational in nature. Like there’s some financial advisors that they’ve done it a certain way for a long time and maybe they’re retiring when their clients are tired and that’s OK that that’s something that that worked out fine for them. Many people are looking for a much more engaged in Financial advisement experience, and different generations are also looking for more. Do-it-yourself kinds of experiences, especially as you have relationships between Sydney, the parent and the children, and the grandchildren. So human displacement and human skills, Human Enablement are what you’re trying to drive from both your own employees and your customers as part of your AI solutions. I’m and in this space probably the second biggest concern is around quality. What if it gets it wrong? What if it’s the wrong answer to this question and that is where I supplied a lot of sort of solace because there’s a lot of ability for us to test the outcomes of their AI solutions and to build confidence around the outcomes that we can validate. In many cases, that’s not the case for our financial advisors. Today they are making decisions based upon their own sort of perspective at the time and if we can guide them to make more make decisions that are optimized, then maybe the outcome for our customers will be optimized as well because they’re very data driven and they’re very prescriptive. And then this last one that comes into play in the financial services industry for sure and particularly coming into play with some other financial scenarios like loans and approvals is bias and this is more about understanding your data and what the data will be used for and using it for ways that are appropriate for the outcomes for different person human groups as well as ways that that data might be used and unintended consequences is super important and something that consider as part of your responsible AI adoption. So all these things need to be part of how you are going down in adoption pattern and if you can’t check everyone in these boxes, you need to pause for a minute and make sure you know how to check the boxes. So as we start to think about AI and finserve use cases, understand that there’s sort of three different domains that might be happening with parallel paths inside of your organization that you might be adopting. You might have commodity AI solutions that are starting to let be leverage for things like next best action or even things like writing an email. You might have semi custom solutions where you’re putting things together. Maybe you’re using something like copilot studio to build solutions or power platform, or in this case, which is probably the primary scenario is you’re building a custom ML scenario, client portfolio optimization. You’re turning data into a decision framework. You’re building a high visibility chatbot that gives very specific answers and the reason why I bring this up is because you have some of these other use cases happening within the the ecosystem. It’s important to note that these kinds of things, especially for financial services, next best action solutions really do require a high degree of persuasion and accuracy. Not that that isn’t obvious to many of you on the phone on the call here, but it is something that really keep in mind that Azure building these it requires very intentional engineering. So let’s go through how this all sort of fits together in the context of next best action. O some of these banking and capital markets scenarios are more about customer support, where we’re going to see next best action really take hold is in a how do I respond to a customer support scenario. And this is something that’s sort of not always thought about in next best action, but it’s certainly one that is important because the unhappy customer is 1 less likely to do effective work with us. That gets connected into the knowledge base scenarios, so things that I need to know to provide a great customer experience. What happens when I have an unhappy customer? What do I do about that? How do I provide them the best Support experience from that path? So customer support is sort of scenario one. Scenario two is about creating a relationship between wealth management teams, the multi generational relationships, the pitch book and action that happens within those teams to create this ongoing relationship with a customer where I’m optimizing their portfolio and aware of what’s happening within their portfolio. So I can drive intelligent insight to make them select the right choices to lead to the right outcomes and all that is tied in with a team of people that are helping them. Maybe some people have very specific expertise within that, that team and one of the challenges can be getting those people to work well together. Actually selecting the action. Understanding who I’m communicating with and next best action can really guide teams to be able to select the right choice, especially based on previous activities that have happened with the same customer or other likely customers that can create great outcomes which relates into insights that you might be gathering from their portfolio and helping to have intelligent conversations with those customers. And then also conversation about risk management and this is about essentially analyzing portfolios, looking for risk, looking for opportunities for your customers to have better outcomes and for enabling your advisors to be able to have intelligent conversations with those customers that really relates into all these things claim automation that’s more of insurance scenario and then fraud detection certainly comes into play, which is less of a next best action conversation and more of a a protection scenario. OK, so in this space I’m seeing two different types of innovation happening. One is a sort of slow, incremental innovation surrounding the work that you’re already doing with them. So you may have a flow that a customer goes through presently and that flow is simply being augmented by AI or even a rules engine to help them to make intelligent choices. And that’s less about disrupting the entire market and more about just helping them through an existing journey. The other scenario might be this starts to pivot toward changing the way the market engages with your customers, and perhaps that’s about letting certain types of groups within your customer set to do more work on their own. And I know that’s already a sort of area that this business space is contending with is how do I enable more of my customers to optimize their own portfolios in the context of choices they can already make? And then how do I enable us to gain more ground on top of what they’re already doing via investment or via advisement? So the disruptive engine can be about usurping the way that many, many advisors have traditionally done business and helping a model that might be more suited to certain groups in the population or ways of people who would prefer investment in certain area or types of income levels to gain outcomes that are specific to them. OK, so let’s hit the the the scenario on customer service 1st and then we’ll go into the sort of financial advisement zone. So let’s say that you have a user who calls in and, or a person who’s being advised and they have a negative experience, or they have to make a change to their portfolio and they’re having a conversation with someone within your team and many organizations. They started to record these calls and they started record them historically. Really low quality call centers just to and almost as like a check Gen general check, but where this is started to go is to coach the person on the call to even make real time changes or real time optimizations to the story that they’re driving with that customer. You can already see this in certain tools, even in PowerPoint with things like presenter, Coach, you’re starting to see this be part of the conversational flow. So we worked with some very large scale organizations around adopting this across many, many, many call centers to be able to understand like what is the quality of their calls and how do we coach our individuals to make better choices when they’re helping a customer to have a better relationship with us. So that might mean an appropriate intro. It might mean information that was provided in a way that the customer accepted. It might mean converting the tone. Did I move the customer from an unhappy state to an ambivalent state to a happy state? Did I have an effective follow up coming out of that call? So next best action sometimes can even mean like what is the next best action inside the current conversation I’m having with the customer real time to lead to a better outcome with them and it’s amazing what’s happening with real time automation right now to enable people to make effective choices. This is also something that of course can be done at scale, so you might find that this is something that could be applied across an analytics basis to know like how all my call centers doing like who’s the best at this, who’s the worst at this? How did I do? And you’re seeing some of this start to happen? Inside of you tools like copilot, you know which are excellent at interpreting something that happened within the context of call. Relating that back to a balcony that had happened as a result of it in it being very clear to understanding even the tone of that happened during the meeting, you can see where it is. Polite, respectful, like this is not something technology could do. Even umm, you know, a year ago this is something that’s coming so fast. In order for us to enable ohh it, it is understanding that’s more akin to the way human thinks about this picture than just like a one or a zero is is whether or not like did I have a particular data point that was in that conversation? No, I’m asking about the tone of the conversation and what. How did the tones start? With what are the tone end with our even things that I can gather based upon insights from calls. I also might then need to understand the action items and pull those action items into how that person then helps that customer. So in the call, maybe we’re prescribing next best action and then maybe we’re also capturing those actions and pulling those into activities at that person might take. And I think this is really sort of low hanging fruit from many organizations because normalizing customer service has become something that’s really tricky for organizations to gain ground in because there’s a diversity of talent, of longevity that they’ve been in the organization, knowledge that they might have, even arming them with the right knowledge is a big step. So I I take this next step, I took a a cash flow analysis. That’s kind of go into some of these financial things, took a quick cash flow analysis from monarch umm, captains to be this is not my finances, but it’s a good example of 1. And I was. I thought this was interesting way for us to kind of start the conversation, which is what is a non obvious conclusion. From this image and UH-2 certain extent, you can gather a lot of that yourselves, right, which is cool. But one thing that we couldn’t do with tech even very recently was have it reason through. If you’ve used the word reason right, if you have it reasoned through a diagram or outcomes in a way which is similar to that of a human. So some things that this started, I sort of pumped this into an AI experience and asked what’s a non obvious conclusion from this image and the image you sent is a screen shot of financial management, software interface etcetera, etcetera. You have steady income managing expenses. Well, you have a high savings rate above the average. This sort of return that back from a general sort of general average spend a large portion of income on services, subscriptions, memberships you may want to review the expenses. So this is like an unprompted view of, hey, you’re a financial advisor, me Anonymous. This what I found from this is you can start to ask questions. You start to ask questions. Start to direct. What are my optimized choices based upon certain scenarios? So maybe the high savings rate is great, but the way that they have optimized that into the different parts to this portfolio are not the right choice. Maybe it’s this idea of like you don’t have any side hustles. And you might want to start diversifying your income streams. All these are choices that you might start leading a customer down, and I think the idea of combining very analytical driven AI, next best action models that are doing prescribed that are doing prediction and prescription with GPT centric models that have this sort of human conclusion, reasoning sort of interface bringing those together to help create a meaningful engagement with a customer either through directly through the customer financial advisor or directly to the customer through a bot type enter type interface or a side loaded interface can be really powerful. You can also see a scenario like this where you might take an analysis from a customer and use that at scale. So for example, did an analysis of a Monte Carlo summary of a customer understanding the upside case, median case, downside case? And here it’s finding that there is 0% successful trial of the couple’s current savings and investment strategy, highly unlikely to feet meet their financial goals. It’s essentially concluding something from a set of data that’s exists in the form in the content, but not necessarily something that it of individual might always know. And what’s interesting about this is you might load up a series of questions, so you might say like I need to know 25 things about a customer support folio before I go into an advising session with them. What does their Monte Carlo analysis say about their portfolio? What is their current net worth? How is it distributed among different resources? Tell me about their existing net income. What’s their current expenses? You might dump all that out into an interface, or it might ask certain questions and almost like de risk, the customer Azure going through the appropriate conditions and then based on that you might even doing it scale what percentage of portfolios have greater than 25 chant percent chance of missing savings goals after they consider the Monte Carlo summary? Now of course you can do that through a traditional model, but that might also miss certain correct they’re try characteristics of the report or aspects of it. So you might combine those two things together to arrive at what an appropriate next step might look like. So how does this all kind of come together? Well, you have a team of people that are working as wealth management teams. They are. They’re working on focused areas. Maybe you have the team member that works on uh, long term financial planning versus another one is doing their operational planning versus another one thinks about retirement like taking money out and it’s difficult for a person to maybe be an expert at all modalities that exist within the picture. One way to think about AI in this context is it is essentially another team member which is kind of odd to think right like, but essentially his goal is to function as another team member. Its goals should be for your financial analysis to individuals to be able to offload these capabilities to that sort of financial analysis, AI team member to do that research like we’ve been talking about and meeting Prep, portfolio analysis, or even prescribed what they’re going to do about it, which is essentially what we’re talking about in next best action. So you might say research this possibility for my customer return this data from My Portfolio. Tell me about recent situations that have happened or what changes should I make and this is truly where you get to next. Best action is what should I do about that based upon our investment philosophy, and that’s where I’ve seen it truly come home for many organizations. Is is not just about like generically. What would you do about this? Like some questions are right, like save more money. You don’t save anything like save more money, right? But there’s a difference between, say, maybe what a just a drop, a name like maybe what? Dave Ramsey might save versus H&R Block. Convert the A and some large financial planning organization. I say right, there’s a there’s a difference between those. Some philosophies you need to determine what your organizational philosophy is. It’s not just up to one person. What’s your organizational philosophy and how do I align that to the prescription associated with a particular wealth management team and allowing that to what my customer’s goals are? So you might have an organizational philosophy, but then maybe there’s specifics you can capture about a customer that allow you to pivot that philosophy based upon their individual goals, and that’s what can help your financial advisors not only to be more effective at helping your customers, but also be more effective at prescribing the right action that leads to good outcomes. Yeah, if if they’re trying to go through a certain period of life, maybe that period of life is just it just objectively optimized by taking a certain action that we happen to know about. So when you think about personalizing those contents, essentially what you have is you have a customer and that customer has historical accounts and behavior data. And we know this because we interact with them as financial advisors, right? So like we have a certain degree of that now. One thing you might also find is that you don’t interact with them enough, so you don’t know enough about them as an individual advisor. So you’re just kind of winging it? That’s not what we want. We wanna optimize the experience so we have this historical data and this would be digital events that have happened with them, what they’ve attended, conversations that we can gather into the picture. This might be like conversations with their financial advisor that we’ve recorded and transcripts that are able to be summarized. So for example, let’s have a, say, a financial advisor has a great conversation with a customer, and they what traditionally would write that down in notes on a piece of paper or in a financial planning document or in CRM or something, right? They capture that somewhere. How effective are is a average financial advisor at taking notes in a standardized way from their customer? Maybe not that great, right? We wanna optimize that process. So what if I simply had call transcription on it and I used the tool like copilot to be able to capture that customer call? And we’re able to suck that into effective call transcripts that could be used for next best action like that would be tremendously more effective than what they potentially did in the past. And they’d be able to optimize their time investments. So even simple things like that can start to invest in the actual data you have to optimize the portfolio for a customer. This might also then move into chats you’ve had with them or what they bought or haven’t bought in the context of your financial products. Or their demographics, but you have to be extremely careful about because that’s where some of that sort of biased situations might come into. But ultimately you have this historical body of data that enters into different drivers in the portfolio, things you want the prices that they’ll buy things at the risk that they’re willing to take, the quality of relationship they want with the financial advisor like is this just a robo situation or do they really want a like super high quality ongoing relationship or maybe different parts of that family want different levels of relationship with us? So all of that feeds into your next best action portfolio that says, based upon these criteria, the data about the person, the behaviors and the risk profile where then feeding that into. Here’s what you should do about it. This is your optimal action based upon this set of things, and that might mean a call like like based upon what we just learned, we need to get in touch with them because there’s problems with their portfolio or there’s actions that need to be taking. Maybe we let it lapse. We let them go. They’re not. They’re not a great customer for us. We just let them move on. Maybe there’s opportunities for them to choose to. Buy something from our portfolio based upon some criteria or we run a promo with them, but even more so like we’re trying to influence them to make great optimal behaviors. We want to help them to make great choices, and this is traditionally like a true ML model, but we’re finding is that that ML models combined with some Gen AI capabilities and other things we know about that customer to be able to provide a great experience. And that’s where I think the Gen AI world is starting to really help help us create a great outcome. This what we can do with static data. Mmos behind the scenes enables us to create great experiences, and that might be in some level of Communications back to that customer around different ways to optimize their portfolio, whether it’s through a person that they’re interacting with or it’s through a AI agent or both that are then working as a team member. We have this idea of like a a wealth management team essentially being and AI team member in the context of the work that you’re doing. OK, so after rut personalizing communications, personalizing customer experiences and action are what is at the root of next best action. So I have a customer. We want them to take action on something based upon improving their Net Promoter score, their conversion, their engagement, the lifetime value of their portfolio and making the right choices. If we don’t take the right actions, what happens is there are more complaints, there’s more people who move on to other advisors. There’s more costs, they have worse lifetime value and ultimately we know those choices like in many cases we we know as a business what a customer can do to make their organization stronger or sorry to make their portfolio stronger. And if we simply had the ability to tell all of our advisors and all our customers to take those actions, we know that our portfolios would be stronger. We know they’d be better off, so this is about leading them down that right path. And AI enables us to do that in a way that is much more tuned to their specific needs then we ever have that variability to do before and it also enables us to be a great partner for the financial advisor to help them to be more efficient as they have to deal with a variety of scenarios. This also can come into play with multi generational relationships. So let’s say you have a relationship with the parent and that is your primary customer that you’re doing doing financial advisory rework with. You don’t have a relationship with the child, or maybe you have with one of the children, and you certainly may not have any. Oh, I spelled that wrong. You may not have any relationship with the grandchild, so this idea of this idea of optimizing the relationship between different levels of the family and using AI to help with the optimization of that. So this may be more online or or app based inner experiences or maybe even this is just getting them into the conversation around financial management and making them part of the portfolio family. But oftentimes, the way that the advisor has worked with the parent is very disalignment to how the child wants to work, and you have to figure out with the AI platform and other tools how to cross that barrier. How to market to the different levels of this group but also create relationships across this group with a cohesive set of tools that helps them be effective and that might mean that this, this team, this wealth management team. Is made up of agent agent, agent may be an AI agent that all work together with the same portfolio, understanding and actions that the child and grandchild might start here and then start to expand to have a full agent as part of their picture or even as a result of relationship with the parent you’re starting to market into and do proactive actions within those children or grandchildren activities. But that’s all driven by next best action. And that’s all driven by us prompting people. If there’s one thing we know about sales in general advisement and generals that sometimes people, there’s really great people that outperform their portfolio and there’s people that don’t and a lot of times it’s about doing the work. It’s about doing the actions that need to be done to be able to provide the right outcomes, and when people do the work a lot of times they make their own luck and that’s what we want to do with our customers. We want to make our own luck by helping them to be successful and all of these kinds of relationships can drive benefit as a result. This also comes into play with risk management, so a lot of risk management happens here in the ad hoc space we’re taking choice, making choices based upon things that we’re reacting to, and sometimes that is the worst choice to make because we are reacting to A to a we have a situation where reacting to it where we want our customers to be is in this proactive and intelligent space. We’re taking choices, making choices specifically because we know they will lead to great outcomes and we’re driving intelligence through storytelling about what could happen if different choices are made. So maybe if someone’s thinking about buying a particular type of property going into this sort of business, adding the side job creating a uh server that they’re considering renting out this house or whatever it is, right? They’re storytelling through different use cases, and we want to help them to make intelligent choices about that and even model their portfolio based upon those choices. That’s another area we’re next best action and AI can provide value to us by helping us to be able to model those out successfully. So this is an example of how this is so. Moving into some things that are like adjacent and related that I think will help you to see how AI is helping drive some very interesting transformation in different Financial spaces. You’re seeing activities that were traditionally very manual in nature and human driven becoming very AI driven. So this is an example of an organization right now that is just gone live in California. That is streamlining the application process for an average savings of about $2000.00 for every loan application that’s going through the process. So essentially cutting $2000 that used to go to the loan officer out of the loan application process by using a chat driven assistant essentially to help the customer through actions in the loan application process and helping them make the best choices through that process. And one of the things that I saw through the ways that they were building that was the number of questions that they asked successfully that a given advisor would never know to ask was really significant. So they essentially built into the tool that combined knowledge of many different individuals that know how to validate these loans before they even get to the hard credit pull. That’s something that you’ve also relates into financial advisory in a sense that like some person might know, this some person might know this. How do we evaluate that and help them all to make great choices? And this is where AI can streamline you if you’ve done your taxes this year, you may have also noted that HR Block and Intuit have launched Advisor Bots alongside with the chat experience of alongside with the tax experience. I think this is really nice way for them to start going live with this capability because it feels very natural as you are filling out your taxes, which is akin to maybe even another financial plan. In a sense, it’s starting to ask questions along the side where you miss something or there’s something you left off the list, and to certain extent it’s rules driven, which many next best actions are. But in this case it’s asking them in a very intuitive way, and it allows you to ask questions about things you might not know the answers to, and directs them toward different self help options before it escalates them to a true customer service scenario. So you can see how in this case they’re building each of these agents right into the experience of them completing a financial process, but then enabling that to also connect into a human experience that they might even have charge for. Once you need to move into that spot, so next best action in this ecosystem is both customer support and driving them through an optimal experience that leads to a better outcome, particularly in the case of like looking for the ductions or other things that are very complicated in the context of the tax code. But they make it really easy in the context of the actual customer experience. So as we close this, I want to do 2 things. I want to address some human first AI strategies that relate to this and be I wanna make sure that we also leave a little bit of extra time at the end for questions. So as we’re finishing this up, make sure you drop any additional questions that you have there in the chat and I’d be happy to answer those as we hit the end of the conversation. Umm, so the first thing I want you to think about is Azure adopting these capabilities and you’re lighting up lighting up commodity capabilities for your advisors or Employees as well as building some of these capabilities that create competitive advantage. The thing we can’t forget about is that this is all impacting people and if we miss that, we’re not only going to miss the chance to drive appropriate adoption, but we’re also going to miss the chance to great get great outcomes and optimize the person experience as a results of this O one thing that you should be thinking about is that across all jobs, not just financial services, but across all jobs, you can expect a significant pivot from creative work, from repetitive work to creative work as a result of the moves that are happening within artificial. Intelligence. And that’s not something that every employee within your organization is prepared for. Many people come into your organization and and they’re very comfortable preparing the financial documents that the they’re gonna review with their customer and talk about in go through in a very repetitive way and maybe they’re not really ready that like they might not have to do all of that activity in the future. And that space is gonna be open the way I I sometimes think about it is when someone has nothing to do. How effective are they in your organization and finding something that’s effective to do and some people are really good at that like they are great in moving from working in the business to on the business in a sense. Many people, however, have not used that skill very often, and they’re used to coming in and doing a role, and that’s where they’re sort of personal self worth. That’s like attached to is that role and this is going to be a pivot even in the most positive leadership environments I’ve seen that someone was positive. Leaders have said one of the hardest things that had to deal with is just The Who moved my cheese moment like the person who is so built up in a technology that they had were using for a while and all of a sudden they built something in AI that replaced it very quickly and did it better. Being able to get on board with that as opposed to being a defensive situation or we’re pivoting their skills. So this move from 80% repetitive work, 20% creative to almost flipping that over on its head to more repetitive work. So sorry to more creative work. So how do we get started? Is it to learn data science curriculum? Is it to get off the ground? You’re gonna have, especially in Financial Services, some data scientists and people that are building these kinds of models, especially in next best action. But it’s not the big change to your staff. The big change to your staff is about Enablement of adoption of tools. So when you think about this in the context of adopting next best action and among your employees, it’s about this idea that everyone deserves an AI assistant. Everyone like you know, sometimes you have an intern. Sometimes you don’t like and some people are really good at working with interns. Why is that? Where the reason that some people are good at is cuz they’re good at delegation, right? They’re good at like taking a set of things and giving them something to do. Like when you have your head kits cleaning the house and they just like they come back to you and you always have a new job, right? You always have something else they could do. Like there’s I have no lack of creativity around coming up with jobs to clean my house. So everyone deserves that AI assistant to take on tasks, do tasks, return them to them in the level of that assistant is continuing to improve. And its capability and then adoption of AI, it’s not necessarily about the adoption of new roles so much as it is about the changing of roles with new skills, which is not limited to technical skills. It’s about reawakening creativity and requires a growth mindset in every person in your organization. So this is like you’re like, well, this is like not a it conversation. This is like a employee HR, conversation, leadership, conversation. That’s exactly why I kind of end with some of these things and the last the last thing before I start to talk about what you can do next is about the roles that exist in this picture. So as you’re preparing some of these solutions, you are going to have a data engineer that is preparing your data for use in the scenario you will have data scientists that build the trusted models that are used that have accuracy and precision and can be reviewed and can be tested and can be evaluated. All of that fits into this picture. You then have an AI engineer that may use foundational models to create experiences within your applications. So like, maybe you have a next best action model that’s been built on a set of AI capabilities, but then you’re layering that into separate experiences. Maybe there’s a customer experience. There’s an advisor experience. This experience, you’re layering that into these different spaces. So AI engineer or developer but where this is going to happen the most is in the AI practitioner who uses this to create value in everyday work. They’re an expert on the business and their job is to enable that activity within your customers and within your organization. So where we go from here? This is what we’ve seen companies do that have been very successful. So first they think about Persona based jobs to be done. If you’ve heard about jobs to be done framework, it’s essentially knowing that like we do, a collection of activities that activities have. That you may sort of source job that needs to be performed like the job. I don’t care about like the preparation of a report. I care about having a great advisement experience with my customer. Their job is to have the great advisement experience, and I either need the report or don’t either report the question is like, how do I have great advisement experience and they’re in that advisement experience. I need to do XY and Z so understand the jobs we’ve done associated with your personas and then enable predictions of information to inform decisions that happen with that job to be done that then prescribe action on that normalized action and then you get to a point where you can analyze single portfolios based upon that one workload and that might mean a specific thing you’re predicting your prescribing for. And then from that get to a point where you can analyze the the general portfolio associated with those same things, you can do it at scale, you can apply it across the entire data set. You can apply it to any customer group of portfolios or think about it in the context of individual actions. But to get there, we have to start with the end in mind, which is that Persona job to be done, and that’s where these conversations effectively start. So as you leave here, I want you to think about a couple of potential next steps. 1st is AI and copilot executive visioning workshops. These we are doing every week. I have 3 executive sessions just this week with the CEO, CFO, CEO. Like these relationships of people who are trying to understand this in the context of their organization, this is something that we do all the time. So executive envisioning workshops as well as driving that into company envisioning workshops that are centric to like teams, IT investment teams, understanding how we can take advantage of this and then we can also do use case validation associated with opportunities to make this real within your organization. So as you believe, make sure you complete the survey. I would love to know if this was useful to you. BI would love to have more conversations and then with that I would love to take some additional questions. So if you want to drop any additional questions you have in the chat, I would be happy to take those as we go. So feel free to put those out there and I will. I will hit them now. See entered the chat. OK. Sounds like you guys are good. I am super appreciative of you spending some time with us this afternoon and I hope this was helpful. Please fill out the form as you close down and I’ll look forward to having filed conversations with each of you about leveraging AI to create next best action systems. Have a great afternoon.