You will leave this session with a better understanding of:
- The AI-Powered Advisor:
Discover how AI tools streamline routine tasks, allowing advisors to focus more on strategic client interactions and personalized advice. - Enhancing Client Relationships:
Learn how AI-driven insights enable advisors to deliver highly personalized, data-driven advice, fostering stronger and more meaningful client relationships. - Power of Data:
Unification and normalization of data is paramount to empowering AI. Centralized digital warehousing is increasingly necessary for firms. - Best Practices for Implementation:
Gain practical insights into successfully integrating AI tools into advisory practices.
Transcription:
Michael Djurdjevic (00:10):
Good morning and welcome to day one of Advise AI here. Originally, we were going to have my colleague, the Vice President of InvestCloud Sterling Perkins up on stage speaking to you all today and sharing with you on this topic, but he lives in Tampa. He's dealing with the hurricane situation there, so as you can understand, we had a last minute change of plans he tagged me in. I have to admit though, when he called me and let me know what I was going to be presenting about, it's more than a little bit excited about having the opportunity to meet you all and talk to you about this myself, because today we're going to be talking about something that's going to change the lives of most of us in this room within the next two to three years, and that's how we can unlock advisor efficiency with AI.
(00:46):
Now, lemme go back. Slide here. I wanted to start because I see a lot of faces I don't recognize in the room to get to know you a little bit better. Simple show of hands, who here has used AI for anything work or personal related within the past month or so? Okay, almost all of us. How about over the past week? Wow. Over the past day. Wow. What I can tell you is that a lot of hands went down over the past day, still a lot up here, but when we come back to this conference next year and we ask that same question, everybody's hands are going to be raised not next year, certainly the year after that. The truth is, all of us who have used these tools, if we use ChatGPT, if you've used Dolly, if you've seen AI generated images, video or asked a prompt of ChatGPT, you know the power of the tool, and you almost intuitively understand that these tools are going to change the way that we work and we do business in the future.
(01:40):
And when I think about the financial services industry and our adoption of AI, it's at a bit of a crossroads because we have technology forward firms that today are training their advisors on how to use the tools, how to get the best value out of them. They're looking at their data, their tech stack, and making sure that it's enabling AI, it's ready for what's coming next, and those firms are the ones that are going to win the most. They're going to see that quantum leap in efficiency and output from their advisors as AI tools continue to get more and more sophisticated, better and better year over year. At this point, almost month over month, it seems right. The good news, what I could tell you is just by being part of this presentation, being in the audience, being here for this conference, all of you're early adopters, you're definitely ahead of the curve.
(02:22):
Now, let me give you a little bit about myself. My name is Michael Djurdjevic. I'm a Sales Engineering Architect with InvestCloud, but I actually started my career 11 years ago as a software engineer in New York. I helped build the first few iterations of the wework.com platform back pre-scan, pre collapse. It was a good time back then, and I moved on to PricewaterhouseCoopers where I was a consultant. I helped build mobile apps for credit companies, networks like Visa and MasterCard, but also credit bureaus, so Equifax, TransUnion, and eventually I took a contract gig with InvestCloud For the past six years, my role at InvestCloud has been almost the same. I sit between our engineering teams and our clients are firms, advisors, financial planners, asset managers, and I asked them, what is it that you expect out of a next generation AI enabled wealth platform? What kinds of capabilities, what kinds of insights?
(03:10):
What do you expect AI to be able to solve for you in the future? With that insight, we were able to build our AI platform we call advisor copilot, but what I'd like to do is tell you that you're going to come away from this conference, this session, certainly knowing three things. One, you're going to understand the importance of centralizing your data with regards to your AI strategy. Two, you're going to have some concrete use cases, some real world examples of how AI can unlock efficiencies for your advisor and for your firm. And three, you're going to have some takeaways as you go back to your offices and think about implementing AI for your own firms, your own day-to-day lives. Now, this is what we're going to cover for the day. I'd like to start again by talking about data. We'll move on to the power of data warehousing and how centralizing your data really unlocks the full capability of a trained AI platform.
(03:54):
We'll look in the future and see the AI powered advisor. What kinds of things is an advisor going to be able to do with the power of AI going forward? After that, we'll talk about strengthening client relationships via user story, where we can look at an advisor servicing their client and how they use AI to better service that client more quickly and produce personalized content and tailored recommendations for them. Finally, again, we're going to end on those best practices, things that you can take back to your firm as you think about building out your own AI strategy. Now, I wanted to start with data because everybody here knows that data is really important. That's obvious. Firms spend millions and millions and dollars of cleaning up their financial data to make sure the reporting is accurate. The inferences they draw from their data is accurate. So we know in this room that we need to have high quality data for anything we want to do, but with AI, it's especially important because the way these tools work as a lot of I'm sure is they sit on top of a massive pile of data, which they iterate over, they study, they analyze, and they're trained on that data to build a model of it.
(04:51):
And from that model, we're able to ask questions like when you go to ChatGPT and you ask a question of it, it is using its inference. It's all of its learned stuff about the data model has been trained on. To give you an answer, in our case, we want to feed our data models, our AI models rather, high quality financial data. We want to give it information about our financial planning, historical investment returns client data from our CRM, and we want it to be able to give us useful analytics that go across all of those systems. So when I say quality data, I'm really talking about unified and normalized data. We want our data in one place, so when we train our AI model against it, it has visibility to all of that ecosystem and it can draw inferences across systems. That's where we see the most value.
(05:33):
The good news for everybody in this room is that even if you have a minimal AI strategy, a couple of KPIs, a couple of reports, some simple things that you've built into your platform, if it sits on top of high quality data, if it's trained against the right data, you're going to have a fantastic outcome. It'll be an efficiency multiplier for your advisors and for your business. So really in our view, from our experience building our platform and talking to our clients, the critical first step if you want to have AI adoption, centralizing your data inside of one place. To that end, what I want to do is show you an exercise that we did with one of our clients. This is a whole lot of stuff here, but this is a real world example of a client that we did a data warehousing exercise for.
(06:10):
You see on the left here, the advisor portal, data sources, and on the right, the client portal data sources in the middle is a data warehouse. Now, your firm, your technology stack doesn't have to look like this. Maybe you're using Snowflake or you have Hadoop in the middle, or you've got a smaller shop and you're running Tableau, whatever the case is. The point is we want all of our sources of data funneling into this one unified data store, so when we train our ai, it's looking across the systems. Why is that? Well, if we look at these individual roles of integrations, right? This really reflects the complexity of what we do in our day-to-day lives, working for financial institutions. I mean, you've got your portfolio data, your investment data accounts, tax lots, transactions of course, but you've also got all sorts of other data that you deal with.
(06:53):
Day-to-Day, you got your streaming market data from your cloud quote or from your Bloomberg. You've got your CRM data, your notes, your analytics on your clients, all that from your Salesforce, your MS Dynamics. You maybe have some pre-calculated performance data. You've definitely got maybe a separate financial planning module and e-money or Money Guide Pro, and when an advisor logs in for the day to do their work, I mean, it kind of looks like this, right? You open up one window for your outlook. You might have another window for your CRM to see what's going on for the day. Another window for your investment returns. If we were to take AI and just apply it on each of these modules independently, train our models against the data sets one at a time, what you're going to find is that that AI can only draw inference about the data that it's been trained against, right?
(07:35):
If I take my CRM module and I apply AI and I tell it to learn about that data, you might be able to tell me that most of my prospects convert on a Tuesday, and maybe that's a useful bit of information from my firm, but it's not exactly the kind of insight that we want to really supercharge our growth. We want to be able to answer complex questions like, what is the profile of an opportunity that ends up investing a lot of money with my firm? Who is that person so I can target them? I can double down on that winning strategy. The key is in centralizing our data inside of one store, we're unlocking the full capability of analytics across the ecosystem. It's really what unlocks the complexity of the questions we want to answer with ai. So as we move forward, I want you to keep in mind that central data warehouse is what's going to enable us to answer those multi-variable complex questions that we need to help grow our business to give advisors some real efficiencies.
(08:25):
Now, if we can accomplish that, right? If we can get all of our data inside of one place, then we're unlocking tremendous capabilities for our advisors. When we worked with RAI platform, what we found is that there's a number of categories, a number of key things that advisors look for that is really supercharged by this centralized data. The first is proactive alerting. Now, this category is someplace that AI can produce alerts that we really can't replicate today due to the complexity of what we want to alert against. This is a simple scenario. Alert me if any of my client's financial plan probability of success drops below 75%, and this is something we can do today. We don't need AI to produce this alert. If you have AWS, you're doing CloudWatch or something, you can set this up, but where this gets interesting is with the power of AI, you might be able to generate an alert that's looking across different data systems.
(09:12):
Maybe you only want that alert for a certain segment of your clients, right? Maybe you only want that alert if the reason for the drop in success was to an outflow of cash from the portfolio or a trading decision that the client made. That kind of complex system alerting is something that we can do with ai incredibly difficult to do today. Another category that we see major opportunity is an advisor intelligence. This is a place where we can give advisors tools and input and analytics that help them shape how they interact with their book of business, what strategies they adopt for bringing a new revenue to the firm. In this case, the example. Give me a list of all at risk clients within my book of business and the last correspondence we've had, I love this. Again, multi-variable, right? That at risk designation is specific to the firm.
(09:56):
Maybe that's a combination of usage analytics. We want the AI to be trained on the usage analytics to be able to understand who's been logging in, who hasn't. Could be email correspondence or messaging, right? Who's been sending emails, who's active. It could also be something like investment outflows or inflows into the portfolio. Has this person moved any money recently? So if we want to be able to generate that kind of report today as an engineer, I can tell you you couldn't pay an analyst or an engineer enough to do that. It would take a whole lot of manual sprawling through the data. AI would be able to generate a report like this nearly instantly. The last opportunity that we see that's major is business intelligence. This is really where firms are going to be able to understand what's winning for them and double down on those approaches.
(10:34):
In this example, we have what client segment is referring the most business over the last 12 months, over the last six weeks. We're flexible here on what the time period is, but the key is we're able to ask questions about what's winning for us. If we can understand that, we can double down on our niche as advisory firms, as wealth firms, we can keep continuing to service those clients in those ways, pushing those financial products. Now, I wanted to go through a specific scenario, kind of a real world use story, so we could talk practically about how AI might help an advisor directly servicing a client. We see here on the left on the top is Michael Chen, who's a mass affluent client. We've got right underneath Sam Smithson, who's an advisor, and then the executive executive management team, rather, Javier Morales for the firm. If you look at this, this is just a timeline of events.
(11:22):
As Michael is going through a life event change, he's having a newborn, right? This is a situation that I'm sure a lot of advisors should have had to deal with. One of your clients calls you and tells you, Hey, I just had a kid. I need help navigating the situation. So week one, on the bottom, we see here that Michael will go about adding his newborn son to the profile. What AI can do for us here is of course, alert us to the event that that's happened, but more importantly, it can guide the advisor as far as what next steps should they take, and when I say guide the advisor, a trained model would be able to tell you exactly what products, what calls, what outreach would make the most sense in this situation, what has worked before for the firm. So when the advisor logs in, they're told, set up a call, talk to 'em about setting up a trust account, maybe get an UGMA account on there, they're given the next steps in a way that they don't even have to think about it.
(12:08):
They know intuitively what to do next. Now, as we move into week two, Sam or rather Michael is actually going to go ahead and open a 5 29 account for the kid. So again, I will alert Sam that that's happened, and in addition to that, it can recommend some next steps about what Sam should do. Maybe the AI system can even generate a proposal for SAM that's already considerate of Michael and his spouse's investment preferences, the current assets they have invested with the firm and also already bakes in their financial plan and is factored into the probability of success. I mean, can you imagine a world where you're in this situation and after hearing that the client had a newborn within that same day or that same week, you're able to turn around a tailored customized proposal that already is included in the financial plan? That's powerful.
(12:51):
That's something that when you go back to Michael, that same week, he's going to come away feeling like a million bucks, right? Because he's got rapid turnaround, he's got a personalized proposal that's bespoke to him, bespoke to his investment preferences, and is already considerate of his financial goals and where he wants to go. Of course, Michael Enthus is going to accept that proposal, and in this scenario, the executive management team will review the new account, we'll approve it, and we're off to the races. This is an example of how at every step of the way AI is guiding Sam, not just alerting her as to what's happening in her book of business, but also giving her some recommended next steps, things that she should do based on objective insight into what's worked for the firm before and what products are the most valuable to sell to this client in this situation.
(13:33):
In addition, that proposal generation is a tool that's going to be massively impactful, massively powerful for advisors being able to generate content like that on the fly. So when we were building RA platform at InvestCloud, we've talked about the journeys, how we can service clients with ai, the different ways that AI can offer efficiencies and tools, both to the advisors and to the firms. I want to talk about some of the experience that we've learned from implementation and from building our own platform, working with clients, deliver that platform. When I talk about implementation best practices here, I'm an engineer. This isn't about technical implementation, right? We're not describing how do you add this to your tech stack? Rather, we're describing what are the key things that you have to do to have an AI platform that people will actually use it, they'll adopt, that'll grow with your firm.
(14:21):
The first thing I want to cover is that less is really more. When we think about AI adoption, it can seem like this massive kind of gargantuan task, truly not that crazy, right? You can start off with just a couple of KPIs, a couple of reports, a couple of prompts that you've written that provide some useful analytics to your advisor. What you'll find is that at advisors use the tool, they're naturally going to come to you and ask you for new things, new ways that they see the tool being useful, new insights that they want, so you really don't have to do too much here. You could start with a small module. It's built on top of that great data. You're going to have a good experience. Next, we encourage you to measure what matters. So as we're talking about the kinds of things we want to provide our advisors, you want to double down on, or rather, you want to identify the KPIs, the reports and metrics that are the most useful to your firm, that offer the most value to advisors.
(15:09):
Again, things like what strategies are winning, what client segments are bringing in the most referrals, what financial products that we sell, lead to the best outcomes. Identify what those questions are, what those analytics are that are going to provide you the most value, and then you're going to be able to measure those and start with those as your pilot for your AI platform. Finally, personalization is key. Anything that you want to release to advisors in the digital space, I mean, the great magic of digital is that it's personalizable. You can log in, you can tailor it to your liking. We released our tool to our advisors. What we found is that we needed to offer them a selection of KPI so that they felt they had some choice and what it was they were doing and how they were interacting with the platform. To that end, we encourage you to make sure that you offer, again, a broad selection of things that the advisor might be able to ask of the platform.
(15:54):
You don't want to open the box completely, but you want to give them enough tools that they feel like they have some control and some input into what's happening. Ultimately, again, when we think about AI, it might seem like this kind of gargantuan task, this just huge thing that's incomprehensible and difficult to imagine implementing like any technology, but it's really not. It comes back to what I said at the beginning. Great data is everything. You can centralize your data if you can get it all in one place while maintained, and you can run some AI models against that. There is really no limit to the level of data-driven insight of data fueled growth that your firm can experience or the kinds of efficiencies that your advisors can unlock. Thank you very much.