Kirsten Chang (00:00:00):

And welcome to today's webcast sponsored by LPL Financial. I'm Kirsten Chang, senior industry analyst at VettaFi. And today we're diving into an incredibly timely discussion all about AI for advisors, separating signal from noise. In this session, we're going to run through the practical advantages of AI when it's fully embedded into workflows, help you find the hidden hours in your day and use them to spend more time with your clients, help you ask the right questions before you trust a tool, including compliance and liability concerns before letting AI even touch client data, and help teach you how to start small without feeling too overwhelmed. Because the fact is AI can be a lot, but the goal is to help you identify one high friction workflow, solve one problem, and build from there.

(00:00:39):

Before we get to all that though, I do want to go over a few housekeeping items with you all. First off, if you have any questions of your own for our speakers, don't be shy, just ask away. You can type your question into the Q&A box at the bottom of your screen. We'll try to get to as many of those as we can before the end of the session.

(00:00:53):

You'll notice a number of modules and widgets. Feel free to move those around, condense or expand any of them to your liking. And for those of you joining us live, some of you already found them, feel free to use those emojis to react in real time to anything you're hearing whenever the mood strikes.

(00:01:06):

You can just use that reaction feature at the bottom of your screen. So feel free to play around with those and have some fun with that. Love to see the hearts and the thumbs up, keep it coming.

(00:01:14):

Quick note, you should be getting an email sometime tomorrow with all the slides you see today, as well as the recording of the full presentation. So don't worry if you miss anything the first time around. You can also find the full deck from today in that resources tab down below, so feel free to check that out. And periodically we'll be asking for your input in the form of live polls.

(00:01:31):

You'll hear a little ding when those polls go out. So keep an ear out for those. Your answers will often help drive the conversation forward. We have a survey we encourage you to fill out anytime throughout the webcast. You can find it right where it says survey at the bottom. So please don't forget to give us your thoughts before you leave today.

(00:01:46):

And lastly, if you're here looking for CE credits, you're in the right place for that too. Just click on that certification icon below and you should see a green check mark up here once you've done everything needed to fulfill those credits.

(00:01:57):

All right. Now allow me to introduce today's presenters, John Stevens, a Senior Vice President of Product Management AI at LPL Financial. John, happy Friday eve to you. Thanks so much for joining us today.

John Stevens (00:02:09):

Thank you so much for having us. It's a pleasure to be here and interacting with everyone on the call.

Kirsten Chang (00:02:15):

Yeah, absolutely. It's great to have you on such a timely topic. And Miller Staten is Head of Product Operations at LPL Financial. Miller, we have you on the line as well.

Miller Staten (00:02:25):

Yep, I'm here. Hello, everybody. Thanks for having me.

Kirsten Chang (00:02:28):

Great. Excellent to have you both. Before we dive in, just start off by telling us a little bit about your backgrounds and what drew you both to the LPL space. Miller, why don't you kick things off?

Miller Staten (00:02:40):

Sure. So I lead product operations at LPL, and my team's main responsibility is to support our product org in delivering new digital experiences for our clients. And I guess why LPL? I joined the firm about three years ago and before that I worked at a few different larger companies like Morgan Stanley and Ernst & Young, and I also spent some time at a couple of smaller startups.

(00:03:05):

And I think what really drew me to LPL was that it seemed like a perfect mix of a larger company that's a little bit more established, but that still had that more FinTech-like innovative mindset. So for me, it was a perfect chance to join a fast moving company that has a really clear and big goal of being the best firm in wealth management. So I'm super proud and excited to be here.

Kirsten Chang (00:03:27):

That's awesome. And John, how about you? Tell us a little bit more about yourself.

John Stevens (00:03:31):

Yeah, it's a pleasure. So I am an Aussie for those in the call, so if I sound like I have a cotton ball in my mouth, that's why. But my long story short is my wife studied abroad and came back with some extra baggage, namely me, about 12 years ago now, and had been in several FinTech and InsurTech firms.

(00:03:48):

I joined LPL not only for all of the cultural reasons that Miller was just articulating, but also because it's in such a fantastic competitive position to take advantage of this new technology wave with AI. It's the system of record for its financial advisors. It has access to advisor data across the full life cycle of the business.

(00:04:11):

And because of that self-clearing broker-dealer and custodian, it really operates as one of the most vertically integrated data ecosystems inside of wealth management. And that structural management, it just has such a, excuse me, structural advantage is so competitive over the horizontal AI companies that we wouldn't be able to replicate the end-to-end workflow orchestration, the way that we serve advisors anywhere else. So for builders of AI products, that data foundation, and then also the amazing advisor workforce and the cultural reasons that Miller articulated are why I joined LPL about eight months ago now.

Kirsten Chang (00:04:48):

Oh, wow. All right. Excellent. Well, I love the enthusiasm. I just want to bring the audience in right off the bat, get a pulse check to gauge our current audience's comfort and familiarity with AI in their everyday workflow, which is the topic du jour.

(00:05:01):

So how would you describe your current use of AI in your practice today? A, not using it all; B, experimenting on my own with some things like ChatGPT, other LLMs; C, using one or two point solutions like note taker or drafting perhaps; or D, it's embedded in your daily workflow and you are very familiar with it.

(00:05:18):

And we'll give everyone a moment to log their answers. And my personal understanding is that AI adoption in wealth management is accelerating, but it's still fairly shallow. So we see some firm leaders who haven't touched it at all, others who couldn't imagine writing a client summary without it. So I'll be very curious to see where our audience lands today, and let's take a look here. So again, how would you describe your current use of AI in your practice?

(00:05:40):

And it looks like survey says, okay, so 45% experimenting on my own. Some of the LLMs out there, ChatGPT, Copilot, Gemini perhaps, 25%, a quarter, saying we're using one or two point solutions for some note taker drafting tool, perhaps some admin tasks. 17, let's say 18%, not using it at all. So definitely some converts in the audience to focus on. And then 12% say that it's pretty embedded in your daily workflow.

(00:06:09):

So I guess, John, on that note, why don't you give us your initial reaction to these results and then kind of tell us what you are seeing as far as advisor adoption of AI so far. I mean, how does this align with the conversations you're having and what are some of the key trends to note?

John Stevens (00:06:24):

Yeah, it's a great question. So on the polling side, this is actually very consistent not only with what we see in the industry, but also what we see within our own advisor workforce inside of LPL. And what we're really starting to see though is a separation amongst those who are adopting it and the types of practices that they offer, which have been particularly interesting. So what we see is in the higher end of the market, advisors are now starting to really lean into AI and embedding it as a part of their workflows and the higher end network clients they solve. And part of the conversations we have with those advisors is that they're so busy trying to make sure that their clients get every single piece of white glove service that they deserve, that they have to figure out how they can make their practices and operations more efficient. And so it's really encouraging to see that these numbers continue to improve over time from an AI adoption perspective.

(00:07:19):

Now, of course, this isn't the easiest trend to jump into and adopt, particularly given we have a regulated environment that we operate within. And so deeply sympathetic and looking forward to continue to figure out how we make adoption of AI easier for our financial advisors as we move the ball forward. In general though, I think the second part of your question is what are we seeing with the wealth management industry and what they're doing with AI?

(00:07:45):

We actually see this in a few different buckets, but the first wave of AI was really focused on these assistants, the note-taking tools, etc. And now there are proprietary advisor assistants that are getting to adoption. There are some leading wirehouses that have measured statements that 98% of their advisor teams are using their assistants. And so that increase in productivity around having that sort of digital twin that sits down with you and completes that work in your day-to-day is really where we see the industry continue to drive forward as well.

(00:08:18):

The other aspect or segment where we see the AI adoption happening is in firms helping advisors think through where should they focus as a part of their clients as well? So this comes in the form of the next best action products that we see, but those next best actions are now starting to move away just from insights, but also shifting into, well, here's how you can draft that email to that client and then make sure that incorporates the tone the advisor has, etc.

(00:08:49):

The third area we see the industry moving is really in the meeting tools and now shifting beyond notes to actually looking across the entire meeting journey because it's where advisors at the highest touch points with their clients and it's where they give their clients confidence in the decisions that they make. And so the preparation for that meeting is critical and it has to span not only CRM notes and emails, but also things like their portfolio performance reviews, market news data, and a huge plethora of data points that they bring in.

(00:09:18):

Then of course there's a note-taking and then they then shift into post-meeting executions. And this is where we see the wealth tech firms leaning in very heavily around end-to-end workflow orchestration, just to take away the nitty-gritty things that makes it difficult for an advisor to build a better relationship with their clients because they're in the back office activities with their assistants versus giving their clients confidence in the decisions they're making. And that's really where we see advisors' strengths being as a part of that relationship building and trust component too. So those are kind of the broad industry trends and outcomes that we're seeing folks moving to as well.

Kirsten Chang (00:09:59):

Yeah, and John, the media landscape, it feels like wherever we look, it's become saturated with stories about FAs and a whole host of other jobs out there being disintermediated by AI. So how real of a threat is this and should advisors be concerned or is the reality perhaps more nuanced than that?

John Stevens (00:10:16):

No, it's a great question. So we actually had a discussion with Anthropic recently where we did a panel with them, and their perspective was that this actually represents an incredible opportunity for advisors. Peter Nolan over at Anthropic actually shared a story about how their internal developers are using wealth management module as a part of Claude. But the key request that they had in terms of feedback was, well, they needed to talk to somebody about this to validate the insights and actions that it has.

(00:10:48):

So even in the most technologically advanced companies with AI, those folks are still looking for somebody to trust. So the way that we see this doing is we share Jensen Huang's perspective, which is people have tasks and they have purpose when it comes to their jobs. So tasks kind of distract us from this purpose. We sort of alluded to that earlier where the purpose of financial advisors is to help build that relationship with the client and give them confidence in the decisions that they make.

(00:11:15):

So that way our job at LPL and other leading firms, and this is what advisors should be expecting from their broker-dealers and other relationship managers, is how do we remove those day-to-day tasks from their day so that way they can focus on the things that matter? I know Miller, you have had some great interactions with advisors around how they're thinking about using that extra free time.

Miller Staten (00:11:40):

Yeah, definitely. And I think we're really seeing this play out in terms of the time unlock in really two main buckets. And the first one is AI is giving advisors time to do more of the things that they already love doing today, like spending time with clients, strategic planning, or for the finance nerds like John and I, doing some more investment analysis.

(00:12:03):

But the second big bucket that we're really excited about is that AI is also unlocking brand new opportunities that maybe weren't available in the past. So some of the more transactional services like tax planning or estate planning, they're starting to become much more automated and commoditized, and with that automation, advisors are actually really well positioned to absorb a lot of these ancillary services to provide more holistic advice to their clients. So for that reason, we actually see it as much more of an opportunity versus a threat.

Kirsten Chang (00:12:36):

Yeah, it makes a lot of sense. And I guess obviously the market is flooded with no shortage of AI tools out there. So how has this shifted the competitive landscape, John, and what does this mean for advisors and their value proposition when all is said and done?

John Stevens (00:12:50):

Yeah, it's a great question. So what we see from the plethora of tools that are available is that the competitive landscape is now democratizing those core skills that were initially specialized in terms of how Miller articulated that to a broader market available to the people. So what we see is that the 30% of capacity unlock in terms of what we're hearing from advisors that they can reallocate that time and effort from opening up new accounts, executing money movement components, to more value added activities as well.

(00:13:22):

One of those value added activities could be expanding the number of services that they offer, but not to what Miller was speaking to, but we also see advisors going deeper with existing clients and using this as a great mechanism to increase the wallet share that they have with those clients as well. And then the organic referral rates we see going up with those advisors who are using the AI platforms to free up that time and have deeper and stronger relationships as well. So it's not really something that you can just kind of plug in and play at the moment. AI will eventually get to that point in time.

(00:13:57):

But to unlock it as an enabler, we know that these folks need to have deep structural changes where their workflows become AI centric. So there's a million tools to pick and choose from, but what we find the best advisors doing is looking at, well, what's core to my practice and the value add that I offer and how I get really good at a few number of things before expanding into too many tools and adopting everything at the end of the day as well.

Kirsten Chang (00:14:22):

It can be very overwhelming. So with that, I want to bring the audience back in for our second poll of the day. What is your biggest barrier to adopting or increasing AI adoption? Is it A, compliance and regulatory uncertainty? B, too many disconnected tools, not sure which to trust, no time to learn or implement, my firm's tech doesn't support it or data is just scattered across systems. We give everyone a few moments to submit their answers.

(00:14:47):

These barriers come up constantly in our conversations with advisors between keeping up with SEC guidelines, vetting tech stacks, and then just finding the extra bandwidth in a day. It feels like kind of moving past that initial testing phase can just be such a heavy lift. I mean, these are some very real friction points across the industry. So with that, let's see what's holding this group back the most in our audience today.

(00:15:08):

What is your biggest barrier to adopting or increasing AI adoption? A couple more seconds to answer and it looks like... Okay, so almost 40% compliance, regulatory uncertainty, lots of that clarity that needs to be found. Almost 30%, let's call it too many disconnected tools, not sure which one to trust, just oversaturated, overwhelming amount of options out there. 14%, no time to learn, and 13%, data is scattered across systems with just about 6% saying their firm's tech doesn't support it. So John, why don't you step in here and give us your take on these results?

John Stevens (00:15:45):

Yeah, and this is why I love having these polls is it's just fascinating to hear what other advisors are thinking in the marketplace. I think too many disconnected tools is something that we hear all the time as a part of it. I think the data is scattered across systems is actually another sort of symptom of that.

[NEW_PARAGRAPH]And what we see AI playing a fantastic role in the future is where if you wanted to become independent, the problem with that is that you had to compromise on choice. If you get choice, that means that you then have to be the integration layer basically across your different systems. And what AI will do a fantastic job is actually integrating those systems so that way the advisor no longer has to focus on those kinds of activities. And that's kind of where the data piece is so important to AI. What you get in is what you get out of it basically as well.

(00:16:36):

Compliance and regulatory uncertainty is something that we actually have been working very closely with the regulators on to figure out what are they thinking and how are they going to shape the AI landscape? So I do expect the uncertainty to continue for some period of time, but what the best firms are doing is actually taking a much more conservative approach than what they particularly have beforehand and are putting in a lot of the controls in place so that way regardless of the compliance and regulatory environment, those firms are set up for success to operate as well. So that's where that conservative control framework becomes incredibly important to enabling the advisors to use AI as well.

Kirsten Chang (00:17:14):

Yeah. Were you surprised at all by the sort of small percentage of people that said they didn't have time to learn it? I thought that number would be a little higher personally.

John Stevens (00:17:23):

No. I have to admit no. I interview two advisors every week, and I hear it all the time from them because everyone's so busy servicing their clients and technology is sort of a small part of their practice today as well. So I actually had anticipated that to be hard. So I'm glad that folks are making the time to learn about these things if that's the case.

Kirsten Chang (00:17:45):

Yeah, absolutely. So Miller, let me turn to you now because we hear a lot about different types of AI. Obviously, it's seeping into every aspect of the economy. Seems like almost no sector is unscathed, I shouldn't say unscathed, untouched by AI. Can you talk us through how AI has evolved from your perspective, and how that has unlocked additional value for advisors?

Miller Staten (00:18:06):

Yeah, of course. So there's obviously been a huge burst of innovation in AI over the last 18 months. And I think that poll that we did earlier was a good example of that where it seems like 80% of people on this call are at least using AI in some ways across their practice. I think if you did a similar poll six months ago, that would've been probably 50%, maybe even less. So I think people are starting to get more acclimated with AI and the tools available, but it seems like there's still some ways to go.

(00:18:36):

We really see the evolution of how AI is providing value to advisors really in three main phases. So if you look at the page, you could see first you had machine learning, and this has been around for a while. It leverages pattern recognition to provide insights and recommendations based on available data. So you could think when Netflix suggests a new show to binge based on what you've watched in the past, that's a pretty good example of machine learning.

(00:19:03):

And what the leading wealth management firms have already been doing is using machine learning for tools like next best action, like John mentioned, to help give advisors nudges when there may be growth opportunities based on certain client behaviors. So an example of that would be if a client makes a large deposit or has a concentrated position, the MBA tool might send a notification to that advisor saying it's probably a good time to reach out. The second evolution came with generative AI, and we know GenAI took a big step forward in terms of the general population with ChatGPT and Claude, but the truth is that many of the leading wealth management firms have already been using generative AI already for things like email drafting and AI driven chatbots.

(00:19:49):

I think the really big and really exciting phase is where everybody seems to be sprinting to right now, which is the agentic AI phase. And agentic AI, for those that don't know it goes beyond just providing answers and writing content, and it can actually take action on people's behalves and complete specific tasks. So this is of course a major change and the leading firms are building agentic solutions both to improve their internal operations and to help provide advisor facing capabilities.

(00:20:20):

Internally, I think one of the good use cases that we could talk about is the product development life cycle. This is one that's very near and dear to my heart. But with agentic AI products that used to take large investments and sometimes armies of people to build can now be built with much, much less money and much higher quality and much faster.

(00:20:43):

And that obviously helps us be more efficient and provide better solutions for our clients. And then I guess just quickly on the advisor facing side, John referenced this earlier, I think a good example of agentic AI is everything that happens after you have a client meeting. So in the past, you may have had to kick off each individual action after a meeting with [inaudible 00:21:06] one by one, whether it be opening or replacing a trade. But with agentic world where your AI tool will be able to summarize your notes, suggest which action that you should take, and all you'll have to do as the advisor is press approve for all those workflows to be complete. So agentic AI, total game changer, and something that we're extremely focused on.

Kirsten Chang (00:21:27):

Yeah, that's definitely sweeping the nation in a big way. So John, bringing you back in here, when we step back from the practical tools and look big picture, what's actually possible today? What does a fully realized AI workbench look like for a modern advisor?

John Stevens (00:21:43):

Yeah, it's a great question. So on what an actual day-to-day experience looks like for financial advisors is really around these five personas that we've articulated down the bottom. So we tend to think in terms of the assistant who sits down and helps the advisor look after their clients and takes care of all the back office activity. We think about the trader, those folks who are really focused on having a hyper personalized investment strategy associated with those folks.

(00:22:15):

And then of course, an investment analyst, those folks from all the financial nerds like Miller and I were joking about earlier, and then people who are much more planning heavy focused. And then of course, not only are you an advisor to clients, but you're also a practice manager. You have to manage your practice and ensure that you're doing the things and the activities that set you up for success to continue growing.

(00:22:36):

And when we think about that benchmark, sorry, excuse me, that bench, in the assistants space, we kind of talked about things like new account opening, account lifecycle management changes. Those types of workflows will end up being abstracted away from you and your assistants. That doesn't mean that there won't be a human involved in the loop. After all, we have to have a human that looks at everything and makes sure it's okay because AI, whilst it's a great tool, is still imperfect.

(00:23:05):

We call it probabilistic versus deterministic. What that really means is that 95% of the time you'll see that it's accurate and ready to go, but there's still going to be an edge case. And that's why the human loop is so critical there as well. So there are systems out there. Oops, sorry, I got some feedback, but there are systems out there that you can expect to then take off those tasks from the day.

(00:23:29):

For the traders, looking at things like, "Hey, what are the tax law implications that I need to think about?" More of having a personal consultant sitting next to you and pulling in the right information, the right time for you, very similar to the investment analyst. That's where the benchmark is what they can expect there as well, bringing in market news stories, etc.

(00:23:47):

For financial planning, the goal of financial planning won't change. I think I saw a question here earlier. It's like, "Hey, why would people still use us?" The problem is that AI is never really going to be able to listen to a conversation and look at somebody understand, "Okay, this is actually what's important to that client, the emotional intelligence."

(00:24:10):

And so that's where the planning side still needs the advisor to provide the inputs of, "Well, we talked about this and then we figured out that this other thing is really what was important to the client, and so we need to make adjustments down the flight." Now, AI will help make those adjustments from a planning perspective and it'll make it easier to understand those inputs, but that's where the human part really still comes to play. Then of course, from a practice management perspective, having a daily feed that articulates, hey, there's this activity that you might have forgotten about from an email a month ago that you need to follow up with or clients have RMDs that need to be processed, etc. Those kinds of nudges is a part of your day and having a feed of those activities is what advisors can expect as well.

(00:24:53):

But there's some core principles that also back up that kind of conversation. Basically, we need to make sure that any of these tools that are delivered are transparent and trustworthy. We have to make sure that the data that's being acted on is referenced so that way the advisor can pull back and say, "Okay, actually, this is not the right thing. I was thinking about this instead. Oh yep, that's exactly correct as well."

[NEW_PARAGRAPH]And that's the role that we play as humans is more of that judgment on the back side of the data there as well. What's great is that AI has the potential to support every role in that practice. Where we're seeing the leading firms focus on is really taking care of that day-to-day activity side for advisors, and then also on the practice management side in terms of those tasks that what make up great service for our financial advisors.

Kirsten Chang (00:25:46):

Yeah, absolutely. We've been talking a lot about AI at a high level and how it can be integrated into advisory practices. But Miller, I want to turn to you to talk to us about what you are doing at LPL. How is LPL putting AI to work for advisors? And if you could, maybe start by digging into the recent announcement around the LPL Latitude.

Miller Staten (00:26:06):

Yeah, sure. So for those that missed it, about two months ago, we did announce our unified technology experience called Latitude. And what Latitude is the culmination of about $2 billion of technology investments that we've made over the last three years that's helped us strengthen our data, modernize our infrastructure, and enhance our capabilities across all of our advisor and investor facing platforms. And with Latitude, we're bringing all this together under one umbrella, and much of this is already available to our advisors today. But what is new about Latitude is that we'll be increasingly embedding our advisor workflows with agentic AI across all the personas that John mentioned on the previous page.

Kirsten Chang (00:26:50):

All right, thanks, Miller. And with that, I want to put up another poll here, a bottom line takeaway question. Does your current firm have a clearly articulated AI roadmap for advisors? A, yes, B, no, or C, not so sure.

(00:27:02):

And while I give them a chance to answer that, why don't I just toss out one of these questions? We're getting a ton of questions from the audience here. Okay, so Greg is asking, "Would we be able to leverage AI tools like ChatGPT and Claude to help audit our website, identify and plan optimization opportunities and improve our social media engagement strategy?"

John Stevens (00:27:23):

Yeah, that's a great question. So for each firm, they need to check with their compliance team, the policy of that firm, of course. At LPL, those tools are available from a public perspective. You can use public information inside of those, but can't use any PAI or client data. And that's really what's key is we make sure our client data is protected and isn't being trained on as well. So there's an official LPL policy, but each person here should check with their firm to make sure they can see what their policy says they can and cannot do.

Kirsten Chang (00:27:59):

All right, all right. And with that, we have the results of this again. Does your current firm have a clearly articulated AI roadmap for advisors? 50% say no. A quarter say not sure. Let's call it a quarter say yes. So kind of split, but mostly a notable portion feels they lack a clear AI roadmap from their firm. So I guess, how should advisors navigate this uncertainty? John, talk to us more about LPL's AI roadmap and how it supports advisors throughout their business life cycle.

John Stevens (00:28:30):

Yeah, of course. So LPL's roadmap, as we mentioned, is laser focused on the assistant. So a practical thing that we have today that's available in pilot and will be coming towards the end of the quarter is the LPL we're calling service and/or knowledge agent. So one of the core challenges is that LPL's platform is really large, and there's lots of resource center content out there to consume.

(00:28:53):

So trying to figure out, "Hey, how to open up an account for this specific situation or move money," you can just ask and get a real-time answer that's going to be specific to your particular circumstances in that. And that response comes back in roughly around a five to 12 second timeframe. So it's a real time experience for our advisors there. So we want to make sure that we're making it easy to do business with LPL first.

(00:29:18):

The next sort of wave of agentic capabilities that are coming, one is around financial planning. One of the top considerations that we hear from our financial advisors is that clients don't actually read financial plans and creating financial plans is really time-consuming and difficult, even though it brings a lot of value to their clients. And so what we're working on is actually delivering the ability to summarize those financial plans and then ask any questions and tailor that to the individual client experience.

(00:29:47):

You can go ahead and you can change what the actual priorities for the client are. It might be tax planning, there might be some cash goal that they need, purchasing a house, etc., and the AI will adjust on the fly. I think creates a meeting summary proposal that makes it very easy for folks to move forward with.

(00:30:04):

When it comes to growth, there are next best actions as a part of that, but it also talks about your advisor growth index as well. These sort of form the foundation of the intelligence capabilities. We're actually going to then have agentic capabilities in the form of an address change in account lifecycle management.

(00:30:23):

So rather than having to update data in many different places, you'll just put it in practically in one space and then it'll make all the updates downstream for you and clarify that piece. So this is just some really practical first versions for it, but next year we see that expanding across all the different backend office services that we have at LPL. So account lifecycle management for beneficiary updates and other items. There will be money movement options that'll be coming in 2027 as well. So we're laser focused on the experience you have with us as a custodian and making that really easy.

(00:30:58):

The other part that we're trying to focus on is really on the data. As you mentioned beforehand, data is so critical and key to these. And so the approach that we're taking is bringing in what we call a domain. So you might think of CRM as a domain or a topic an area and saturating that intelligence inside of Cyan, the flagship AI product that we have. So you can then ask any questions in real time about that particular data domain as well.

(00:31:23):

So that's the approach that we're taking from just a practical day-to-day item piece, but then also how do we make this a digital twin for you to interact with as well? That'll eventually culminate into a feed that'll help you drive your business in the way that you'd like your business to function as well. And so the goal here is to make it really easy to do business with LPL so that way you spend more time focused on your clients versus on the backend systems.

Kirsten Chang (00:31:49):

Got it. Thanks for that, John. And with that, I want to bring up one more final poll for you all. This is that hidden hours question we kind of mentioned earlier.

(00:31:56):

If AI gave you back five hours a week, where would they go? A, more prospecting and growth, more lead gen; B, deeper planning work for existing clients; C, more client meetings; or D, back to my personal life, more of that work-life balance here. And we'll give everyone a moment to lock in their choice.

(00:32:12):

Time is the ultimate currency for any practice. Five hours a week is a game changer. That's effectively a full extra month of capacity every year. So it'll be really interesting to see whether this group leans more toward business growth or buying back personal time. John, do you have any guesses here? Not to put you on the spot.

John Stevens (00:32:28):

Well, I know what I would be doing, which is back to my personal life.

Kirsten Chang (00:32:32):

Okay, there we go. All right, so let's see what the-

John Stevens (00:32:32):

The hope is... Go ahead, sorry.

Kirsten Chang (00:32:37):

Yeah. No, no, no, please, of course. Miller, how about you?

Miller Staten (00:32:42):

I don't see a deeper financial investment analysis on here, or that would be my pick, but I guess more client meetings.

Kirsten Chang (00:32:50):

Yeah. Or maybe deeper planning work for existing clients could kind of fall into that as well. All right, so if AI gave you back five hours a week, where would they go? Let's see what our advisor audience says.

(00:33:00):

Okay, so about a third say back to their personal life, same as you, John. And then almost a third, let's call it more prospecting and growth, more lead generation, a little over a quarter percent, deeper planning work for existing clients, getting more in the weeds there, and then 10% or so more client meetings.

(00:33:16):

So why don't you go ahead and give us a quick reaction to that, John? Any of this surprise you or does that line up with what you'd expect?

John Stevens (00:33:24):

I have to admit the deeper planning work for actually more client meetings being down the bottom is very interesting. It's sort of counters to some of the conversations that I've been having with advisors, but I think deeper planning work for existing clients probably ends up meaning that you do have maybe more of those client meetings or longer client meetings in some ways. So I think it's a really interesting poll, but I'm also very happy to see advisors prioritizing their personal life as well, that's for sure.

Kirsten Chang (00:33:53):

Yes, absolutely. And just looking back at those earlier poll results on compliance hurdles being the biggest barrier to entry, that kind of hits the core dilemma advisors face today. To get those rich tailor-made outputs from AI, the system needs context, but the moment you feed that context into a model, privacy and data security immediately enter the conversation. So let's talk about how to navigate that balance safely. John, what in your mind is the biggest risk to privacy and security today when it comes to AI, and what safeguards are maybe in place to protect advisors?

John Stevens (00:34:27):

Yeah, it's a really good consideration. I'm glad that advisors are worrying about this item. So the way that we see it is that they're actually... It's just PII, so it's your client's data and information. You don't want that information getting out there. And there are really four protection layers that advisors should look for from any provider that they have.

(00:34:46):

So one is a contractual agreement that their firm or they on a personal basis have with the LLM provider. Generally speaking, these firms have sort of an opt-out approach of training on the data and you want to make sure that's turned off by default, and there's no language in there that lets the provider potentially get around some of those terms. And that's really the most important play there.

(00:35:12):

The second is that you really want to have a data protection layer as well. And that data protection layer can come in two different forms. So the first one is data masking and tokenization. Basically what that means is reducing out the key specific items that are actually identifying their clients, but also have maybe other medical records or financial information that could be trained on and used externally.

(00:35:38):

The next piece though is when we get to the agentic side, is that agents need to have some sort of identity access management protocols. You don't want your agents being able to access somebody else's clients or being able to go and escalate its permissions and doing things that it's not authorized to do. So identity access management for the non-humans is really important for the agentic world as well. On top of course, your own identity access management rules for that data.

(00:36:08):

Finally, we have two secondary layers we typically see. So one is redaction and hydration. So basically we want to be able to feel like you're using PII with the system, but you want that system to effectively redact and identify any potential PII fields and then remove them whilst enabling you to have that LLM-like experience. And of course, rehydration basically means that what you see as an advisor is just the underlying information that you input, but the LLM never actually gets to see that.

(00:36:36):

And then finally, you want to have a firewall as a last sort of mechanism there as well. I know this is getting a little technical, but these are really important for advisors to understand. The firewall's like a last resort that blocks any of those pieces too.

[NEW_PARAGRAPH]So just to recap, there's sort of three key pieces. One is a contractual commitment. The second is data protections in the form of traditional sort of data measures. And the third is the redaction of hydration and then a firewall associated with that information too.

Kirsten Chang (00:37:05):

I haven't heard the term hydration used in this context. Can you just expand a little bit?

John Stevens (00:37:10):

No, of course. Rehydration is just the process of taking information that was redacted. So for example, let's say, that the advisor put a client's name in there with some other context like an email address. When we redact that, what the LLM sees is like a placeholder.

(00:37:27):

It might say name, one, for example, and there's no way to identify that with a person. But then when they receive it on the backend, it would need to then provide the actual name back to the clients of the client so that way they know that. But there's two different systems. One rehydrates it and keeps it separate from the LLM, so you have a separation of concerns there.

Kirsten Chang (00:37:48):

Okay, understood. Thank you for that. All right. And then Miller, why don't you give us some of the key takeaways here?

Miller Staten (00:37:57):

Yeah, happy to. So if you were going to boil it down just to, I guess, four big takeaways, the first one would be that there's uncompromising focus on data security and protection. I think especially with agentic AI, and some of this sounds like sci-fi, but if your firm is building agentic AI solutions, you should be asking what safeguards are in place to make sure that your data is protected.

(00:38:24):

The second takeaway is that AI automates manual activities and creates new opportunities. And I think the more that you can use AI and leverage it to unlock the time savings in your day to start to do some of those new activities, the better. Three is that human judgment is your unique value prop.

(00:38:43):

So one thing that AI can absolutely not replace is your ability to build relationships with clients and use the judgment that you spent years creating within your own careers. And then four is just streamlining workflows. And I think firms more and more are going to be leveraging AI and agentic AI to streamline the workflows through the entire life cycle of the advisor experience.

Kirsten Chang (00:39:07):

All right, excellent. Thanks so much to John and Miller for fielding all these questions. Before we get to Q&A from the audience, just want to quickly go over these disclosure slide here. AI strategies require thoughtful implementation, appropriate oversight, and alignment with each practice's needs.

(00:39:21):

If you're curious about what embedded AI could unlock in your practice, LPL has a number of resources available for you to check out. You can just scan the QR code or visit those resources to learn more up on the screen. And while we wait for questions to keep rolling in, I just want to take a moment to read this disclaimer at the bottom as well.

(00:39:36):

This material has been created and designed for licensed financial professionals. VettaFi and LPL Financial are separate entities. All securities and advisory services are offered through LPL, a registered investment advisor and broker-dealer, which is a member of both FINRA and SIPC. Tracking numbers is listed on the screen.

(00:39:52):

All right, thanks for bearing with me there. Now let's get to some audience questions here. I guess, Miller, let me toss the first one over to you. Steve is saying, "I'm a solo advisor, not tech-savvy, and I barely have time for what's on my plate now. So where do I realistically start?" That's a great question.

Miller Staten (00:40:11):

It's a really good question and curious on John's thoughts on this too, but what I'd say is to start small and just to treat it like any other investment you would in your regular life. So the more you invest your time into learning more and using the new AI tools in the short term, the more time savings that you're going to have in the long term. And I think an easy place to start is probably to look at where you find yourself spending the most time out of your day on administrative tasks. And that could be drafting emails or prepping for meetings and then see what AI tools that your firm offers that can help you reduce the amount of time that it takes for you to do those tasks. But John, I'm curious if you have a different thought.

John Stevens (00:40:50):

No, I think you articulated well. One strategy that we find that works fantastic is most advisors are members of focus groups or study groups that they partner with. Having that accountability partner to learn with you is a great way to make sure you're actually dedicating the time because it's all about the choices that we make in terms of our focus areas.

Kirsten Chang (00:41:15):

Great. And John, you mentioned a number of tasks AI can help with obviously. Which specific tasks though would you say will deliver the fastest time savings and measurable ROI? We have a few questions about that.

John Stevens (00:41:27):

Yeah, so we have quite a bit of data on that. What we see is the way we think about it is the frequency of those tasks and then of course just the size and number of people that are impacted by them. So the reason why note-taking took off is because it's actually such a frequent item. It's something that advisors do once a day.

(00:41:45):

We do think that emails in terms of helping understand the emails that clients are sending, etc., and responding to those emails will be the next sort of big wave of adoption based off the amount of time that it takes. We typically hear advisors sharing that's between 30 to 40% of their day. It's not just the drafting of the email, it's well, what's all the information that has to go into that response for an email? Another area that we see the specific tasks kicking off would be around account maintenance activities and planning is where AI will play a really strong role to help automate some of the not so fun parts of that so you can focus on the fun parts.

Kirsten Chang (00:42:25):

That's right. And John, sticking with you, what governance structure do you see as effective for ongoing sort of model oversight and compliance? Someone's asking, how do I stay compliant using AI?

John Stevens (00:42:37):

Yeah, so I would lean on our AI governance team here for this answer, but we have a re-certification process on an annual basis where we re-certify the agent end to end for those that we build internally. However, every three days we have a process around drift and monitoring where we check to see is the AI still doing what it said it was going to do and has the same amount of quality. Now that's like an enterprise system and that's what you should expect from any enterprise that you work and partner with.

(00:43:11):

For an individual advisor, it's really about making sure that you are checking the day-to-day pieces and using the industry best standards and practices as well. And I would partner or reach out to your compliance team to help stay on top of that. But that's where the human loop comes into play and just the day-to-day of like, "Hey, if I did use AI," checking, "Does this information actually make sense?" Before using that information is really the best way to stay compliant because advisors are the first and the last line of defense when it comes to these tools.

Kirsten Chang (00:43:43):

That's right. A question from Brian, "How do you ensure AI outputs are auditable and that human judgment remains central?"

John Stevens (00:43:51):

Yeah, this is another great question. It sounds like people have been talking to our AI governance team at LPL. But the way that we do this is that every action that has any sort of impact is kept in a specific log and is retained depending on the type of action and what our FINRA and SEC requirements are for a period of time. So what that means is that there's an untouchable record that's created for those materials. So that's how they stay audible.

(00:44:20):

From the [inaudible 00:44:21] remains central, any workflow process that we create, we end up forcing the advisor, which is the right thing to do, to say confirm before it actually goes off and executes something. Or if it's like an intelligent kind of Q&A, we want to make sure, is this the right information? And so we actually proactively ask those questions as a part of those workflow and/or intelligence Q&A experiences. That's the best way to ensure that the human judgment remains central.

Kirsten Chang (00:44:52):

Got it. And Miller, let me toss this one to you. Kevin's asking, "What change management tactics work best to overcome no time to learn or too many tools barriers?"

Miller Staten (00:45:03):

No, that's another good question. And I think the change management of AI is almost as important, if not more important than the actual development of AI. And I think it's from an advisor perspective, just making sure that you're actually carving out time in your day to learn the new tools, to know what's available. From the actual firm perspective, I think it's extremely important that there's dedicated initiatives where you're actually building materials, you're building collateral, and you have that cohesive go-to-market strategy when you do build new products and services that are AI related so that you can actually get the adoption and the outcomes that you're trying to achieve.

Kirsten Chang (00:45:42):

A couple questions about compliance approval. So Blake is asking, "Is there a reference spot that I can search in LPL that explains what is compliance approved for AI use? My mental barrier is compliance."

John Stevens (00:45:53):

Yeah, that's a great question. We do have an official policy document that should be available for advisors at LPL. We can certainly follow up and make sure that we link that to you.

Kirsten Chang (00:46:04):

Okay, great. Roy is asking, "I see many examples of 'AI slop' in the world. How do I know whether AI is going to actually save me time or cost me time in having to correct its actions or suggestions?" That's a good question.

John Stevens (00:46:18):

It is. And it's sort of the question that Miller and I both tackle with not only for advisors, but also internally in the way that we interact as well amongst our internal employees at LPL. In terms of knowing whether AI is going to actually save you time, the best way to look at that are actually the reviews of the products that are out there and what folks are saying about them. It's sort of true for Airbnbs, Ubers, etc. Those ratings actually end up being a really fantastic way to trust. The wisdom of crowds are a fantastic process to lean on there as well.

(00:46:55):

And then I would do sort of micro experiments as a part of this. When I say micro experiments, don't sort of jump all in into something straight away. What we find works really well is if people try something, have an experience, and then shift away from that if it's not working out for them because every practice is unique and we want to make sure that we're tailoring that experience for those practices.

Kirsten Chang (00:47:17):

Miller, no, go ahead.

Miller Staten (00:47:19):

Maybe I can just add to that super quick. And not to oversell it, but that's one of the things that we, John and I, are really excited about with Cyan, our new AI product, is that it's actually being built on top of the foundation and the data that we've spent years and years creating. Because as I think everybody on this call knows, the AI tools that you use are only as good and as powerful as the data that is feeding those tools. So if you're using a tool that does not have access to the same data, there's a lot more likelihood that there could be hallucinations and the output might not be as good.

John Stevens (00:47:55):

That's a really good point, Miller.

Kirsten Chang (00:47:55):

On that note with Cyan, I guess we're getting question from Ernie. "Sorry if I missed this, but will Cyan be able to act as a meeting assistant from a recording perspective, i.e., Jump or Zocks?"

John Stevens (00:48:07):

So the initial version of Cyan won't have the meeting assistant capabilities inside that in the initial version. From what our focus is at the moment is doing the things that those assistant tools don't offer right now. And we think of ourselves from a vertical perspective.

(00:48:29):

So our unique value proposition and our defensible mode is that we are the best custodian for you to partner with a broker-dealer. And so that means how you interact with us is really where we want to make sure that we shine for you to deliver differentiated value. But Miller, I don't know if you have any thoughts on that that you'd like to add.

Miller Staten (00:48:50):

No, I totally agree. I think you nailed it.

Kirsten Chang (00:48:55):

So Teresa's asking, "How do you know which AI agent is reliable?" She says, "The new AI spell check for my phone is a nightmare. Used to guess correctly. Now it's wrong every time. Spend too much time correcting it."

John Stevens (00:49:09):

Yeah, that's a really good question, and a great example. So in terms of how do you know which AI agent's reliable, we actually go through a very deep and rigorous eval process. And so the outcomes of those evals are something that we're considering sharing and making sure folks understand and work with them on that.

(00:49:31):

So what you really want to see from a provider is what kind of tests and experimentation did they run to validate that the AI works and are they being transparent around those too? If they're not being transparent, then that introduces a little bit more uncertainty and skepticism that we would advise folks have. Miller, I'm not sure if you have any sort of feedback or thoughts from a third party AI tool perspective.

Miller Staten (00:49:55):

No. And I think from the user perspective, exactly what it sounds like you did is the way to know. You just have to experiment with it and see the output that it gives you. It's pretty obvious, I think, when you're using a tool that does not have access to good data because the answers just won't make sense. So I think just using the tools and leveraging it as much as possible to understand what is and is not actually giving you high quality output is really important.

Kirsten Chang (00:50:25):

Kevin's asking about prompts, which we know are sort of the heart of AI and a lot of the LLMs that we used every day. "So the more my firm uses AI, the more I realize the importance of effective prompting. Do you have any recommendations for teaching employees how to write prompts more effectively such as courses, certifications, or other training resources?"

John Stevens (00:50:42):

Yeah. And this is actually probably the biggest sort of concern that we had from advisors at Focus, which was, well, how do I get my employees motivated about that and how are we mobilizing the workforce to be ready to use AI? This specific area, Claude, ChatGPT have really great guidelines and principles that you can use and adopt for that. So they have publicly available resources that you can use.

(00:51:08):

The other piece is the best practice that I've found on a personal basis is asking the LLM to take a concept and then create the prompt, the world's best prompt for that particular thing in that circumstance. The tool's actually getting very good at helping you write their own prompts as well. That's sort of the great part about these products is that if you have a question about is this reliable, how do you do it? You can always just ask them and they'll respond usually with good content for that. Greatness is in the agency of others and training employees is going to be a key area for advisors to focus on.

Kirsten Chang (00:51:46):

For sure.

Miller Staten (00:51:47):

I think you're talking to the right person because John actually built the prompt library for our entire company. And I think that the knowledge sharing is really, really important. Even internally with our own product teams, John hosted a training to walk through what prompts work really, really well for product development and how to actually tweak and refine the prompts that already exist. So I think it's about just making sure that you're connecting with the right people within your firm that are already using these tools really well and have prompts that might be really good for their specific situation.

Kirsten Chang (00:52:21):

That's a great point. I guess Miller or John, Michael's asking, "What is the most popular feature of your offering today among advisors?"

Miller Staten (00:52:30):

So John, curious your thoughts too, but I think the most popular feature today is definitely our next best action tool. And I would probably say two actually. Next best action tool, which does give insights to advisors on certain client triggers.

(00:52:46):

And then the second being we have a generative AI search functionality within our resource center, which is our central knowledge repository, where in the past you used to have to sift through links and try to find answers to the right information, where now you can actually just use the generative AI search functionality to get answers right in real time on the top of the page. So I think in terms of usability, that's probably what's used most often. I expect that to change once we roll out some of the more new and improved SIEM use cases, but curious what your thoughts are, John.

John Stevens (00:53:20):

No, you're absolutely correct. The data from our sort of monthly active usage metrics is pretty indicative that those two tools are the most popular ones that are being used by advisors.

Kirsten Chang (00:53:32):

We have a question about that prompt library. Do we have access to that prompt library in the resource center or where can we direct our advisors to find that when it's available? [inaudible 00:53:44]

John Stevens (00:53:43):

No, it's a great question. It's actually an internal prompt library for product management and engineering, but we are in talks with the resource center and legal and compliance about how can we provide some really good prompts and also our own sort of version of, "Hey, here's how to prompt to that." Unfortunately, I don't have a timeline in terms of when that kind of functionality will be available inside the resource center, but we will be keeping you up to date as that moves forward, that's for sure.

Kirsten Chang (00:54:12):

Sounds like it'll be a popular tool once it's available. Derek is asking, "What are your thoughts on building an internal AI system using local models instead of cloud-based?"

John Stevens (00:54:23):

Oh, that's a really technical question. It's great. So I think there's a couple of different ways to answer that. So maybe I'll answer from an LPL perspective, which is the reason we use a mix of tools, we're primarily an Amazon Bedrock shop, and what that means is we have access to lots of different models and tools.

(00:54:47):

From a legal compliance perspective, they have a favorability towards more of the frontier models, given that we have better contractual terms with those because contracts are a really important piece of protection for those. So as long as you have the contractual pieces in place, if there is a model that you can use locally that is fit for purpose and you go through the rigorous testing processes, etc., and you don't use any PII and you follow your program's policy and firm's policies there, those local models actually end up being really interesting outcomes there. They're much cheaper to use, etc., etc., and they can also specialize too. So we are actively exploring that space, but right now we're mostly focused on the frontier models because the contractual arrangements that we have.

Kirsten Chang (00:55:37):

Okay. And we'd like a follow-up on that email drafting you were talking about. John, Jeff says, "Can you have him explain more about AI helping with email drafting and responding? Is that basically Copilot or is there another solution available to advisors?"

John Stevens (00:55:50):

Yeah, so obviously Copilot is available as a part of that. The experience that we've found with Copilot though is that it doesn't have the broader enterprise context. And so what we're focusing on towards the end of next year and timelines are to be, well, as a part of that email, there might be a reference to a data point inside of our data foundation via Latitude that might need to be referenced.

[NEW_PARAGRAPH]And so the goal is to take a look at the intent of that email from the client, resolve the household context and provide you as the advisor, "Here are the key data points and topics of conversation for you to have as a follow-up for that email response." So nothing available as yet, but it is an area that we're actively exploring and very excited by.

(00:56:41):

There are some other available third party tools. They're not approved, I don't think yet, on the LPL platform. So that's a part of the Affinity Program. Please take a look at the Affinity Program for the tools that are approved and available that might have some email capabilities for you there, but that's something that we're really interested in as well.

Kirsten Chang (00:57:01):

Great. Thanks for that. We have a follow-up question about privacy and security. "I understand using AI for public data, but does LPL have a dedicated AI platform to use with PII, being able to use agents and such within LPL's AI platform?"

John Stevens (00:57:15):

Yeah, so that's actually exactly what Cyan is. So those four layers of protections that we were talking about earlier are actually been built into Cyan to make it feel like you can use PII with it, but in reality, there's no PI that's actually going towards the LLM. And so that's the whole purpose of why we built Cyan, was to enable that kind of functionality for our advisors because it was one of the number one requests that we had. Miller, I don't know if you have any other context you'd like to add to that one.

Miller Staten (00:57:45):

No, I think you nailed it. With agentic AI, I think it gets a little bit more specific where some of the capabilities that we are building is to really understand just like we would for if it were human, one, who's acting, so what is the agent actually doing and does that agent have a unique ID that we can actually track the audit trail of exactly what that agent is doing? Two is keeping agents inside the line. So making sure that there's a very defined scope for every single agent that we build so that we know exactly what that agent is and is not supposed to be doing. And then three is I think it's the fail-safe, so the ability to turn it off and monitoring processes to understand that if an agent does start doing something that we don't want it to do, we could easily flip the switch and turn it off and see what's going on.

Kirsten Chang (00:58:37):

Got it. Thank you for that. Follow up question from Derek. "Is the secure browser rollout a part of the future Cyan usage requirement?"

John Stevens (00:58:45):

We don't have that formally as far as I understand as a requirement today, although that may be a change in the future coming down. Miller, I don't know if you have any context on that one in particular yet, but we can certainly follow up, Derek.

Miller Staten (00:59:03):

Yeah, I don't think it's a requirement, but let us follow up just to make sure we don't mislead you.

Kirsten Chang (00:59:09):

Okay. We do have a question that I think speaks to a lot of what people run into with AI, which is sort of this existential crisis. James is asking, "Why would a client keep using our services if AI is going to do it all?"

John Stevens (00:59:20):

Yeah, it's a really good question. So AI isn't going to do it all. There's been quite a few different technology waves that have sort of threatened the financial advisor, and let's take the robo-advisor platform as an example. It's not necessarily AI, but it's all automated service.

(00:59:40):

And what's been really interesting from a data-driven perspective is that the number of assets and the growth of those assets in the wealth management space separate and distinct from the robo platforms has actually outpaced the growth of those robo platforms in relative terms. Our core thesis about why is that people want to have something or someone that they can talk to and trust. There's not really that emotional intelligence that you'll receive as a part of AI.

(01:00:13):

And then also AI is not going to be able to replicate everything that you do for many, many, many years for us to worry about. And so that's the core thesis in terms of why I joined LPL, is that it's best positioned to do the things that humans don't really enjoy doing and free them up to have that conversation. Miller, I know that you've got some really strong thoughts on this too though.

Miller Staten (01:00:37):

Yeah, I think it's just like we mentioned earlier, it's that trust factor that AI just cannot replicate. And I know John mentioned that we did the fireside chat with Anthropic, and one of the really interesting things that they said was one of the biggest searches that the Anthropic engineers had within their own internal version of Claude was to try to find financial advisors. So even the people that are at the absolute forefront of building the technology are still searching for that human in the loop advice. So I think that should be a good sign that we're in a pretty good spot.

Kirsten Chang (01:01:11):

Yeah. And I do think emotional intelligence goes a long way as well with connecting with your clients. Final question, since we have a minute to go, I guess I'll start with John, Miller, feel free to chime in, but which agentic AI workflows, whether it be onboarding, planning, trading, etc., should be first and foremost in advisor's minds to deploy above all else? What's the number one priority here if you had to choose?

John Stevens (01:01:34):

Yeah, if we had to choose, so one I'll be really annoying and say it depends on the practice of course. So for example, if you're a planning practice, then you should really be focused on the planning side. At LPL, we're laser focused on those sort of key custodial verticals, so money movement, account maintenance, new account opening.

(01:01:56):

We're not picking one necessarily. We're actually trying to make progressive progress on all of those areas as well. Miller, I'd love to get your take though too. If you had to choose one, which one would you choose?

Miller Staten (01:02:09):

Choosing one is so difficult, but I'd prioritize the workflows that save the most time. And I think they probably won't be the ones that you typically expect or the ones that, as John would say, have the most sizzle factor. But I think with Cyan Assistant, for an example, we're prioritizing account maintenance as the first workflows that we are addressing. And those are things like updating a client address or a client last name, which might seem small, but if you add up all the times that advisors need to do that throughout the day and throughout the week, it does actually equate to a big time savings.

Kirsten Chang (01:02:48):

You have it. That does it for today's webcast. Thanks so much to John and Miller for joining us. And thank you to everyone here for attending.

(01:02:53):

To those who submitted questions, if we were unable to get to them, there were plenty. Someone from the LPL team will personally circle back with you. As a reminder, full recording of this webcast will be available starting tomorrow, so keep an eye out for those emails in your inbox. Thanks again to everyone. Have a great Thursday and weekend ahead. Thanks, guys. Take care.

Miller Staten (01:03:10):

Thank you.