Featuring:
Shaun Tucker
Director of Practice Management
Brock Sutton
Head of Emerging Client Capabilities
Nate Angelo
Chief Executive Officer of Composition Wealth
Shaun Tucker: Hello, and welcome to the PracticeLab webinar series. I'm your host, Shaun Tucker. I'm the director of practice management here at Capital Group, and I just want to thank everyone for joining us today. It's going to be great to get together for this discussion. Our topic today, Using AI in your practice: Your questions answered.
And we're going to answer as many of them as we possibly can. So we're going to dive into this all-important topic.
And first, let me just share a stat from our recent survey on advisor AI adoption. And the stat is 80% of advisors are using AI in one way, shape, or form. That's a big number, right? But where and how they're using it in their practices varies a great deal. And with that said, I think there is a growing recognition that we're in a critical moment in time right now where your decision around the relationship you have to this technology, and the degree to which you integrate it in your business is going to be a key determinant to your success going forward. And that may have felt like a bold statement a year ago when we first started talking about AI on this webinar series, but we don't think it is anymore. And so we're going to get into that in today's discussion.
There are so many questions that have surfaced around AI. How do I get started? How can I catch up if I'm behind? What role does it play across every aspect of my business? What limits does it have, and how will I know what they are? And then really importantly, what's the role that people need to play as a complement to this technology to help you maintain your brand, your personalized services, and also that connection that's so important with your clients to ensure differentiation in your practice? So these are great questions, and we know that you're going to have many more. The good news is that we have a great panel today that's here to answer your questions and provide unique perspective.
And with that in mind, we really encourage you to engage with us today, ask a lot of questions throughout the discussion. We're going to do our best to answer them, and if we don't get to yours during the session, rest assured we will follow up with answers after this event.
So let me clear some housekeeping real quick, and then we're going to get into introducing our panel. If you look in the upper right of your webinar screen, you're going to find everything that you need for today, including the Q&A tab, as well as slides for the event. We're offering CE credit for IWI today. And we also want to remind you that a replay of the webinar will be live in about two weeks, so look for that to come out, and please do share it with your colleagues and friends.
Okay. With that, I'm going to introduce our panelists, and I'm going to begin with our friend, Nate Angelo. Nate is joining us from Seattle, Washington.
Nate is the chief executive officer of Composition Wealth, a fast-growing, $11 billion and counting RIA. Before being named the CEO, Nate was the firm's head of wealth management.
And prior to his time at Composition Wealth, he had several senior leadership roles at RBC Wealth Management. Nate is a seasoned professional with 24 years of industry experience, and importantly, he's been a champion of AI and he sees it as an unlock for his team and his business. So, Nate, welcome. Appreciate your time today.
Nate Angelo: Thank you, Shaun. It's great to be here.
Shaun Tucker: Yeah. Great to have you. My other panelist is Brock Sutton. We're welcoming him back to the program. He's my colleague. He is the head of emerging client capabilities and a practice management technology consultant at Capital Group. That's a fancy sentence to say that Brock spends all of his time analyzing the impacts of technology on business and translating it into guidance on how to use it to grow and scale your practice. So Brock's been in the industry 14 years, at Capital for eight. And while he's calling in from New York City, our New York office, he's based here in Los Angeles with me.
All right. Let's jump into the conversation. Again, I want to just stress that the goal today is to answer as many questions around implementing AI in your business as possible, so please don't be shy. And I also want to mention from a format perspective, instead of us having a conversation and waiting until the end to take your questions, we wanted to make this much more interactive today, so please do lean in. Don't be shy. Submit your questions in the Q&A tab, and we're going to try to address them along the way at the end of each of the three sections of our conversation.
Okay. So with that, I'm going to set the stage with our first series of questions for Brock and Nate, and I'm going to keep it pretty broad. And the question is, why is AI so important, and why did you feel compelled to get into it? And, Brock, I'm going to start with you. Tell us your story. Obviously you decided to dive into this several years ago and really go deep. You saw something absolutely critical going on here with AI. Tell us your story. You have such a passion around this. Why did you get into AI, and why is it so important?
Brock Sutton: Absolutely. Thanks, Shaun, and excited to be here today. So for me, I think my interest has always really been in technology, but specifically how technology impacts different business dynamics and business models. So that's kind of the approach that I've always taken to my work. And I think for me, I really got focused on this in around late 2020, early 2021. It was shortly after GPT-3 had just came out, and you could see from GPT-1 to GPT-2 to GPT-3 this progression of what I would call modern-day AI. And if you squinted, what you could see was what was coming today and what was coming in the future. So for me, that's when I got really excited because I said, "Wow, this has a lot of big implications." People that I was talking to were obviously discussing this. People that I was listening to were talking about it. So that's what kind of turned me onto it.
And then really shortly after that, I also got a little bit anxious, and I said, "Wow, with the pace of this technology and where it's headed, a lot of the things that I'm kind of doing on a day-to-day basis and I'm defining as my value prop at Capital Group, it seems like this technology might be better than me at some of those things." So I said, "How do I make sure that I stay economically valuable?" And at the same time, I actually had started working with you, Shaun, and your team and saw what we were doing on the practice management side around business consulting with advisors and our clients, and I said, "Hey, this is too good of an opportunity. It makes too much sense. We are all as knowledge workers going to have to deal with this 10, 20-year transition within AI, so if we're going to be the partner of choice to our clients, it seems like a great space for us to lean into." So that's a little bit around how we got here today.
Shaun Tucker: No, it's terrific, and it's a great story. And I love that comment, "Stay economically valuable."
I should mention that as a part of our Pathways to Growth study that we do on what drives growth and success among the top advisors out there, technology, tech stack integration continues to be one of those skills and behaviors that really does drive growth.
And I think you just saw that AI is starting to really consume that, so there's that piece, but also AI is starting to play a role in every aspect of your business, and we're going to definitely get into that. But maybe, Nate, here's a chance to bring you into the conversation. You're running a very large and successful RIA. Tell us how you've adopted AI in your firm, and why did you feel it was so important to do that?
Nate Angelo: Yeah, great question, and thank you again for the opportunity to be here, Shaun, and it's great to be with you, Brock. What I will say out front is I am not the technologist that Brock is, and so even if I squinted several years ago, I did not see where we are today coming. And so for me, it was a couple years ago starting to think about AI and starting to see just the practical applications and asking myself, "How am I going to overcome my own fear of not understanding or fully appreciating what this technology can both do today and where it's going?"
And the second part of that then was as this continued to evolve ... I think historically we'd seen technology evolve in longer time lapses, so you'd see a software application come to bear and you'd go, "Oh, wow, we're going to watch this evolve over years." And we quickly started to see things happening in months and weeks in progression. And so for me, really I started to get fearful that if we don't take some sort of action and we put aside the idea that we are not proficient in it, we will just fall further and further behind, and so we had to take a leap. And so the strategic component of taking that leap was twofold. One was engaging outside consultants to give us a better perspective of the landscape of what all is out there, how do you begin to think about it, how do you put a framework around all of AI's different capabilities before you could even think about a place to start?
And so then that led to our entry point, Shaun, which was, how do we do something that exposes our entire firm to AI and some of its capabilities, recognizing that with 150 people, there's a lot of different views on AI, there's different understandings of AI? Some people are using it in their daily life. To Brock's point, some people are fearful of what maybe AI is going to do. And so we actually entered by transforming what would have been in the old days, if you will, the intranet site, our operating system as a firm, and we really moved to an AI-enabled platform for which you launch all of your applications when you show up every day and you open your computer, you are in an AI-enabled operating system. And so our entire business now launches from an AI-enabled platform. That was our entry point, and from there, we've built.
But the important part for me was not just to get a certain subset of the firm exposed. It was really to try to help everybody understand that this is going to impact all of us, and it's our job to do this in such a way that it's impacting us and our clients for good.
Shaun Tucker: That's an amazingly encouraging story, Nate, so thank you. So you felt behind, both of you saw disruption on the horizon, and then you were proactive about jumping into it. I love the encouragement, Nate. You got help, right? You engaged outside consultants, right? You kind of laid out the entirety of your business, and then you went full scale into implementing the operating model, if you will. And we're going to get into kind of the how of all of that. But the keyword for me was just disruption and then proactivity.
Brock, let's come back to you. How does that resonate with you? As you hear Nate's story and as you've come to know Nate's story, does that align with what you're seeing from other practices in your consultations? What advice would you give practices who aren't using AI yet on how to get going and why this is so important?
Brock Sutton: Yeah. I mean, I think the thing that we have to realize and I think it's apparent in Nate's story is this is going to impact everyone. But the other thing I think is financial services generally, we are at the very beginning. Shaun, you mentioned that 80% number, and again, within our data, our proprietary research that we do from Capital Group around how advisors are leveraging AI, we see that 80% number that's using it on that consistent basis. But I think what's really interesting is if you go and you have a conversation with most teams, and I would say Nate and Composition Wealth are probably ahead of most teams, but I think if you go and have that conversation, they're really doing just the basics still. So 80% are using it, but it's still really basic. It's content creation. It's summarization. It's a Google Search replacement. That's how they're kind of using AI.
And I think this slide highlights some of the reasons that folks maybe aren't quite comfortable with it yet. One, and we can dig into this in the questions if we'd like, but the compliance and regulatory risk. This gets brought up all the time, so this is definitely something especially our industry is concerned about. And the other one, what I see is a lot of folks that just aren't completely comfortable using the tools overall. They don't know all the features. They're not comfortable. They don't know what to type into the chat. And so we see our financial services industry just really at the very beginning.
The other thing I will say, though, one of the reasons to use it is what we do see in our data as well is that the larger the team, the larger the book of business that they are managing, the more likely they are to leverage AI. So you are seeing teams really start to lean in and start to adopt it, as Nate just discussed, and then we're seeing that more penetration within those actual larger teams.
So I think one of the reasons to do this is this is going to impact everyone. We couldn't have said in the 1980s, "Computers? Pass. I think I'm not going to try to learn that. I don't need to learn computers." And to me, this is the same kind of thing, where everyone needs to learn it and lean in. But then the other thing, your competitors are leveraging this technology. If you're not, other people are. And again, I think with where the technology's at today, it's so economically valuable that you need to start to explore and start to apply it to your business to gain that competitive edge.
Shaun Tucker: Yeah. Great opening comments by both of you. Thank you for the thoughtfulness. Disruption, right? There's an opportunity to accelerate your progress by implementing. Everybody's getting into this, but at varying stages, and we're sort of scratching the surface on the potential here and its impact.
So let's get into the agenda. That was a great just opening remarks. We're going to get into three sections today of our conversation. So number one, where do I start? That's going to include a lot of positive encouragement for those that don't feel like they have yet and are wondering, "Can I catch up?" Absolutely you can get caught up. Let's get going. And so hopefully we'll give you encouragement to do so. Number two, how do I integrate AI within my team? And number three, how do I think more broadly about building an AI strategy across my business? So those are the three things that we're going to cover with the conversation today.
So let's dive into part one, right? Where do I get started? Brock, over to you. We get asked this a lot. What advice would you give on that front? How do we begin?
Brock Sutton: One of the big problems that I see when working with teams is this analysis paralysis. And I think what this points to is that AI is a general purpose technology, meaning that you can apply it across your entire business. But I think this chart here that we're seeing on the screen is really helpful because, again, this is from some of our latest data on how your peers are leveraging that technology. So if 80% are leveraging it on a consistent basis, this is actually where they're leveraging it.
And when I look at this chart, there's really two things that stand out. The first thing is the two big blue bars on the left, so administrative duties, and my guess is is that if we polled folks on this call, this would be their first guess around where it would be applied. And then the other one is on the investment research and selection side, and I would say this has really taken off recently, I would say post-December 2025. That's when the capabilities got really good within this place and that's when we saw a lot more folks, a lot more financial services companies and advisors start to imply that investment research within their existing process.
So administrative duties, what does that look like? I think there's three areas that people typically leverage this technology around. One would be something like client onboarding, so think of those repeatable tasks that you have with heavy bottlenecks where you're trying to kind of accelerate. There's a lot of information that's coming over from the client. There's a lot of forms to fill out. We've seen a lot of folks try to apply it there. The other one that I would say would be our-
A lot of folks try to apply it there. The other one that I would say would be around meeting prep. So whether it's a prospect or an existing client, taking data from all these different sources, aggregating that data together and then coming up with an agenda, coming up with a report that you can then share with the client, and then meeting follow-up would be probably the big three in the administrative duty side. On the investment research side, it's much broader, much more specific to each team. So some of the things that we see there are automating your due diligence for different asset managers. We're starting to see some things around portfolio risk. We're starting to see some things around fund screening. So there's a handful of different use cases within that area, but again, this is fairly recent because of that post-December 2025 kind of evolution of AI where we're starting to see advisors implement it.
The other thing that I think you see in this chart is this massive long tail of gray bars. And again, I think what this points to is that this is a general purpose technology. And what I mean by that is you can apply this across your entire business. And so what matters most is for you as a team, as an organization, to identify those bottlenecks and prioritize those. What are the things that if you could do 10 times faster, it would unlock doubling your revenue? And those are the types of questions to answer. So again, it's a general purpose technology. So one of the important questions to answer is what are your bottlenecks? What are your pain points? A few to just highlight here, report and newsletter production. So if you're not using it on the content creation side, whether that's a newsletter, whether that's a website, that's some low-hanging fruit.
We're seeing more and more folks start to apply this to financial planning, tax planning, estate planning. Again, because some of the capabilities have gotten really good in those categories, we've started to see higher adoption. And then the last one is really in client acquisition. So we're seeing folks not just build their growth strategies with this, but actually implement their growth strategy. So actually go to market, what types of events should they be at?
What COIs (center of influence) should they be engaging with? And then actually building out the content. So again, I think two things here, administrative duties and investment research, those are great places to start. But I think the most important question for folks is what are your bottlenecks, and start to apply AI there.
Shaun Tucker: Fantastic. Rich answer. Really appreciate that. Nate, you have to be smiling because when I think about some of the early stuff that you and your teams have done, Brock just checked a lot of those off. And then I think about the stuff you're leaning into going forward. Love this concept of identifying the bottlenecks. That's a great place to start. We are scratching the surface on this. So Nate, over to you. How have you approached implementing this on your team? And maybe two nuances here that I'd love to get into is obstacles that you ran into as you implemented, but then obviously, where are you seeing clear benefits and success?
Nate Angelo: Yeah, agree. I mean, Brock captured that so well in terms of laying out the framework for all the various areas. There's an opportunity for AI to engage. And there's two very high level points that I would love to address that shaped our thinking. Integration has been talked about for years across the tech stack, and I think AI has the ability to actually truly make that become a reality.
As we all know living in the RIA space as close to integration that some of these disparate systems have claimed to deliver, APIs (application programming interfaces) are not shared equally. It can be challenging. We're finding that AI can actually help us bring together some of those disparate systems. I'll talk about that in a moment. The second thing that we really anchored our heads around, and this is old lingo from the investment world, but was a bit of a core satellite approach to how we think about deployment of AI.
And so maybe it's not deploying AI equally across all these different areas that Brock articulated, but maybe there's some areas that are very core to the business where we get our greatest benefit out of the gates and then over time we begin to think about where do we implement this more in a satellite approach. And so this question of how did we do this? Number one, we engaged an outside consultant that is tied to our industry as an industry leader and we addressed one of the first things Brock said, and that was, is our data lake set up in such a way that our data is secure before we begin to think about overlaying AI into our data, and all that comes with that tied to SEC compliance and where the industry's going.
Second, we actually went to industries that are further ahead than wealth management and picked the brains of AI leaders in those areas to better understand their deployment.
And so where we landed was take all of our process experts that live inside of our organization, deploy Claude broadly at an enterprise level, and then we hired an outside expert to partner with our internal process experts to start deploying. And we went about this exactly how Brock acknowledged it upfront. We went for onboarding, meeting prep, meeting summaries, meeting follow up, and all in the spirit of trying to begin to create more efficiency for our service teams and our advisor teams to spend more time on that human engagement element, both with existing clients, but then also on the growth edge. And the last thing I'll just say here, Shaun, to your specific questions about obstacles, we learned very quickly that you can get things deployed fast, but adoption is imperative. And we missed the boat on having the proper training and ramp-up period either through pilots where we're learning as we go and then deploying more broadly and then having those ongoing training and one-on-one sessions with our larger teams to make sure that implementation is happening at the rate at which we're deploying.
And so we missed that, to be candid, on the front end. We've gotten much better about it now and we see how imperative it is to our success. We can go as fast as we want, but if we don't get adoption, we're in trouble. And then the last thing I'll say just from a success perspective, we are literally now for the first time having people say, "Wow, I have more time." And now we're trying to help them manage that excess time in the most high value add ways. And so it's been great to see people have this aha moment and it come to life for them that this isn't a threat if done right. It truly is an enabled partner, i.e. an agent supporting them and running the business.
Shaun Tucker: That's fantastic. I love the focus on the core business and they're really systematizing things. The training piece, that's huge. I think probably a lot of folks have questions around who did the training. Did you hire someone to come in that is on a permanent staff member or are you outsourcing to those consultants? Maybe they did it. I don't know if you want to answer that real quick, and then Brock, I'm going to come back to you.
Nate Angelo: Yeah, great question. So we are leveraging both external parties who are in whatever that AI system is or technology that we're deploying, and then they pair with our advisor practice management team for the ongoing implementation. And so typically we'll leverage them for deployment, high-level delivery. They'll be on Zoom with a large group of our folks and then we go to small group follow-up sessions to actually get the application adopted into their workflows and into their roles and responsibilities based on team structure as they serve their clients. So it's a combination, Shaun, of leveraging internal and external resources.
Shaun Tucker: No, that's a great answer and I think very encouraging to probably most who are hearing and wondering about that. Brock, let me come back to you, and let's talk about tools. So how should we be thinking about tools as you get going here? I think the classic build it or buy it conundrum is there. What's the cost benefit analysis? How should we be thinking about this?
Brock Sutton: Yeah, and I do want to build a little bit on what Nate was saying because I just think it's so interesting of this idea of OpEx (operating expense) versus CapEx (capital expenditure).
So it's, "Hey, do you spend money and do you spend time actually investing in the technology or do you use it out of the box?" And so what we're seeing more and more firms do is focus on the OpEx side. So they're taking the out of the box solutions and applying those to the business. And if we look at that framework for how we're seeing other teams leverage this technology, so if you look on the right here, we get this question all the time around what tools, what technology should I be leveraging? And so where we want to end up is this core versus satellite approach to the technology. And Nate kind of talked about this a little bit.
It sounds like they're using Enterprise Claude, which we have seen a lot of folks use within this space. At the same time though, we've got kind of these satellite tools. And so I do want to just explain this so folks can have a better understanding of the framework that we're seeing more and more firms take as they approach this technology and applying this technology to their business. So on the left side here, we have the AI tech stack. I just want to briefly explain the top three layers because they map perfectly to that core versus satellite approach. So first up, we have the model layer, and the way to think about the model layer is that's the layer that actually creates and even exceeds human-like intelligence. There's two things we have to know about that layer. One is there's really three or so competitors that are at the bleeding edge. It's Google, it's OpenAI, and it's Anthropic. Every other month, what seems, is one company is declared the winner, the other two are dead, and then three months later we get to a little bit of a different place. The other thing is that layer is moving incredibly fast. And so even in these slides, I think what we have listed is it's doubling every 4.7 months. And what we actually see now is it's doubling every four months. So this technology's moving incredibly fast. And when I say doubling, what I mean by that is the amount of replacement human work that these tools can accurately do is doubling every four months. So it's moving incredibly fast. Now, how you get access to those tools is via something like a platform. And so the way to think about a platform is a platform is something that you can apply across your entire business.
So business management, investment management, client management, you can be analyzing a portfolio one minute and you can be creating a marketing campaign the next. It can really handle all of these cognitively diverse type tasks. And that platform/model layer, that's what makes up the core. And so really what we're seeing is we're seeing four kind of major platforms and core technologies being leveraged by financial advisor firms. So that's going to be Microsoft Copilot, enterprise ChatGPT/Codex, enterprise Gemini/antigravity from Google, and Enterprise Claude/Claude Code. So that's kind of the core of the foundation and this is what Nate was talking about when he was mentioning that core versus satellite approach. On the other end of the spectrum then are your applications, and they may market themselves as AI companies, but they're much more like software companies using AI to solve really specific problems. So the value here is ease of use, but the downside is lack of flexibility.
And so these make up the satellite. They're really good for solving immediate problems, bottlenecks that you guys have as a firm today. So the number one tool we see there are the note takers, things like Jump, Zocks, Microsoft Teams, Zoom AI all fit within that category.
We also see some of the CRM (customer relationship management) tools. So some of the AI features in different CRMs, folks are starting to experiment with those. And then the last thing I would say is some of the AI features within financial planning. So again, we're seeing this core versus satellite approach. And then the last thing I'll just mention is there is a tension within this marketplace. Because that underlying model layer is moving so incredibly fast, what we're seeing over time is it's starting to eat that application layer. So I think there's a broad question of what is the role of productivity software in a post-AI world.
I think you've seen some of this play out in some of the markets as well with different chatter or multiples and things like that, but I think this is an open question. And so I think there is this tension in the marketplace where as these models continue to get better and as you're able to run more and more of your business on your core, how does that complement or how does that work with this approach? So as folks are thinking about this technology, I would encourage them to take this core versus satellite approach as Nate has.
Shaun Tucker: Yeah. And I think there's really good advice there for folks that may feel overwhelmed here is I think the steer here is focus more on OpEx than CapEx, as you say there, Brock, is leverage the industry experts. We're not going to build or compete with that, but I think it's about get to fluency, think about systems, identify your bottlenecks, look at your core business, implement right, and then spend your time innovating based on the tools that exist currently and find meaningful applications that create real scale in your business as a result. So I think that's the steer. Let's open up to questions. So as promised, we're going to get interactive here. A whole bunch have come through in the chat, which I really appreciate. Maybe just we'll start with a question around ... First question is, “Are there any prompts to avoid? If yes, please discuss.” Other question is around the role of AI agents and how should we customize them. Maybe I'll start with those two and see if there's a response from one or both of you.
Brock Sutton: I'm happy to take the prompt one. I'll just give folks kind of two tips, and this will maybe help with folks' usage. One is if you're using these core tools, make sure you're using the right model. We still see a lot of folks are not using the model selector in these tools. What the model selector allows you to do is select a better level of intelligence. So originally, November 2022, we had the original, excuse me, ChatGPT moment. And the way to think about this is this is that instant model in some of these core tools. It's like a PhD intern, a PhD across every discipline in human history, so they're very bright. You go ahead and you ask them a question and they just respond to it immediately off the top of their head. So there's really only three things that you should be doing with this kind of instant type model.
It's going to be content creation, summarization, and fact retrieval. Those are the three things. Anything that is cognitively more intensive, you want to make sure that you're using this reasoning type model. You're going to get much more value out of the product there, because again, that's going to be a big breakthrough. It's that same PhD intern, but now you're asking them a question, they're taking it away for a week, they're breaking it down step by step and they're coming back to you with an answer. So make sure you are using the increased level of intelligence on these core tools.
The other thing I'll say is, again, I think people struggle with the blinking cursor. And so one of the flows I want people to use is to tell the tool what you're trying to do, have it ask you questions, answer those questions, and then have it write the prompt or it create the skill. So what this would look like is to say, "Hey, I'm trying to automate my client meeting prep process. Here's an attached document of what a good output looks like as well as the typical inputs."
“Based on that, what questions can I answer so that you can help me write the prompt to automate this process or skill?” It'll then come back with 10 to 20 questions. What then you'll do is you'll respond to those questions, you'll get all of that institutional knowledge in your team's head down on paper and then you'll say, “answer those questions” and say, “based on that, can you now write the prompt for me to help me with this process?” It's not going to be perfect, but that'll get you much farther than where you would go if you just typed in a few paragraphs on your own.
So I would maybe encourage folks to do that as they're leveraging these tools day-to-day.
Nate Angelo: Shaun, I'm happy to quickly comment on the agent side if you want.
Shaun Tucker: Please. Yeah. And hey, Nate, if I could, there's a lot of questions about where are you using agents. So maybe if you could comment on, “Is it around portfolio management? Is it around financial planning?”
And just I'll throw in another question that came up is, “How do you account for potential hallucinations? Does that keep you up at night?” So sorry, piling on.
Nate Angelo: Yeah. We could do a whole webinar on those three variations of that question. So let me start with the first part of that on the agent side. So for us, we are partnering either with an external what I would call expert in coding and developing agents, or Brock listed off a variety of applications where in which those firms actually have dedicated consultants that will help work with you to create your agents that are going to assist you. And so in those two cases, as I mentioned earlier, we're bringing the expertise in terms of the process and knowing every step of that process and we're anchoring our expertise there bringing that to our outside partners who are then helping us create the agent. And then jointly we continue to work through iteration after iteration with those agents till they get to a point of proficiency that we feel comfortable to do an initial deployment, which is a fairly controlled environment, i.e. a pilot, where we're giving them a limited advisor set in an example of maybe onboarding clients or some client service processes where it's limited in scope and then we iterate again and then we go to full deployment across the firm. And so that's everything from our note taking, meeting agenda creating, follow-up templates to what we're doing for onboarding, what we're doing for basic ongoing service needs of clients. So for us, that agentic partnership and creation of agents is tied to our process ownership outside consultant that helps us, marries our insights together to create the agent behavior we're looking for.
The last part of your question around hallucinations, I think hallucination's never been a positive word and I think to this day it's not a positive word as we think about behavior of either humans or agents. I think this is for us on the agent side is why we have multiple iterations taking place before we get to a full deployment. We've been in a good place so far doesn't mean it won't happen. So certainly it's always on our mind, but we're doing everything we can to be preventative in that way.
Shaun Tucker: Well, I want to make sure to just summarize what I've heard from you, Nate, which is a very deliberate systematic application of this technology. You laid out your business, you analyzed the core components, you understood the bottlenecks, you hired outside expert, you brought process expertise and then you went through and created pilots and control groups iterating your way there, testing and learning, refining. And that's how you can systematically implement a technology for scale and impact. And I think that's an important thing to stress.
Great questions. I'm going to keep us rolling along and we're going to embed some of these questions in the dialogue. But let's move to part two, and it's around team integration and making AI work across your team. We've already kind of just started getting into this, but big question for you, Brock, which is just how do you position your team for success in an increasingly AI driven world?
Brock Sutton: Yeah. And I do just want to pile back onto what Nate was saying, because I think it's so important, which is evaluations should be part of your agent deployment process where you are having a systematic process for evaluating these agents on a consistent basis until they're accurate. So I love what Nate said there. On this front, this is really interesting because I took my own advice here, but I think the way to think about this is technology always has the same impact. Technology always commoditizes something. So it takes the price of some goods or some service and it drops that down to zero or near zero and the beneficiaries of that are the complements. So you want to position you, your team and your organization as a complement to any net new technology. So another way to say this is to say, "Hey, we need to stay as close to this technology as we can. We need to use it in the areas it's great and understand those areas, but then at the same time we need to upskill in those areas where it's weak and that's how we become more valuable."
And I think the best example of this is the inflation deflation chart that we've all seen and we're all comfortable with from the year 2000 on. So if you look at this inflation deflation chart, it has goods and services on the top end that are more expensive today, inflation adjusted. And on the bottom side, it has things that are cheaper, inflation adjusted. So on the top you've got things like healthcare, things like education, childcare, housing. We all know those cost more today than they do in the year 2000. And then on the other end of the spectrum, what we have is essentially goods and services that could be shipped across the ocean in a box and their ancillary services.
So you ask yourself and you go, "Well, what happened?" Well, what happened was in 2001, that was when China joined the World Trade Organization. So you had this additional country of capability that was coming unlocked that had joined the rest of the world. And basically anything that they were able to touch, anything they were able to mass produce, that became deflationary. And at the other end of the spectrum, all of that excess capital then chased those scarce resources.
So I actually think this is a perfect analogy for what we're going through today. So t-zero is 2026, right? We've all seen these CapEx numbers, 750 trillion or so, or billion, excuse me, in 2026 alone spent on these data center build-outs. We have a country of knowledge workers that's coming unlocked to the rest of the world. Your question that you have to answer is, “What are those knowledge workers going to be great at?” And then how do we make sure that we outsource them there? How do we make sure that we apply them in those areas they're great? And at the same time, how do we understand their weaknesses so that we can upskill in those areas and we can make sure our team, our firm, our organization becomes more valuable in a post AI world.
So obviously what you're seeing on the screen here is just a hypothetical, but I think this is a good framework that we can use as we're kind of developing with this technology and thinking about where we should be applying it and then where should we be upskilling as humans.
Shaun Tucker: Yeah. I think this is the central question. Where does AI play a role across your business and where do humans engage more? And Nate, you're passionate about this, but talk to us about what you've done at the team level and what are the functions that got commoditized for you that AI is going to serve a role? Because we've mentioned a few, but I know you have many more. And then where are you leaning in more on the human side and where are you stressing your team's value add, if you will, with clients?
Nate Angelo: Yeah. Building on the theme of this idea of commodities versus complements. The way we're talking about it internally is that this isn't a binary world. It's very much moving to a hybrid world where you have people and agents beginning to work in this hybrid world where they're going to coexist and they're going to augment one another or help one another be better at what they do. I think for a long time we have been in what has been deemed as a services business, but most of the work that we do behind the scenes, whether that is account paperwork to open an account, account documents tied to a private investment, all of these things really are not adding a meaningful value to the life of our client. And so we want to find a way to streamline those efforts in as quick a way but thorough, accurate as possible through the work of agents and/or systems so that our people, to your point, can be redeployed in areas that I always talk about as being referable.
In my view, our world is really about that client advisor experience and what is it that we can do every day to enhance that client advisor experience in ways that create referable moments? Well, how do you do that? Well, people want to be seen, heard, and known. They want a sense of belonging. They want a transformational relationship. That doesn't come by getting the paperwork right. And so for us, it's allowing our people to let go with an open-handed sort of mindset that we can let go of the paperwork, we can let go of the meeting agendas and allow AI to drive that work so that we can then redeploy our human capital against humans.
And I don't think the handwritten note, I don't think coffee sitting in someone's living room has had more value than it does today in a world where a lot of our folks that currently have wealth are asking their own big questions about AI. And so I think the more we can lean into those experiences and time with clients and prospects and COIs, the better we all are in terms of retaining clients, growing the business, attracting more clients, delivering more impact while we continue to invest further in AI around all of the logistics behind the client relationship, if you will.
Shaun Tucker: Yeah, I think that's incredibly powerful. Brock, any comments on what Nate just shared?
Brock Sutton: Well, we've got no video here, but luckily we still have the audio. But yeah, I mean, I think that's exactly right. And I think it's just about rethinking what work looks like. And so if you think about a job, it's essentially a collection of a hundred or a thousand different tasks. And it's very unlikely that AI is going to take away all hundred or all thousand tasks, but there might be some of those that it's better at. And so how do you lean into those other categories and kind of upskill in those areas?
And then I think another good way to frame this is the Industrial Revolution. So if you look at the percentage of folks that were in farming, and I know a lot of people have used this example, pre-Industrial Revolution, what is it? 80, 90%. And then the Industrial Revolution came along and we just got so much more productive in those areas. And then human labor found new things, new ways to be economically valuable. So I really think the human element, in person connection, the relationship building component, communication, I think those types of things are going to matter more in a digitally abundant world. So I think that's 100% right.
And also this is fluid. No one knows the exact answer here. No one knows exactly what's going to happen and what it's going to look like. And that's why you should kind of always be evaluating, always be testing your priors and kind of willing to say, "Hey, I was wrong." And then pivot based on that because we're all going through this together, but I think those are all helpful frameworks to use. I think Nate's really approaching it with his team the right way.
Nate Angelo: Shaun, can I add on one little comment to that? Do you mind?
Nate Angelo: So I think we're all seeking this framework and trying to put something that can feel esoteric or too big to conquer. And so one of the things that we've been doing a ton of reading in this space, but there's this idea of AI transformation and everybody has different lingo and models, but one that we have really anchored around is this idea of this four levels. So AI as a thought partner,
AI as an assistant, AI as a teammate, and then AI as the system. And I think, I don't know if it's 12 months, if it's 24 or 36 months, we get to a place where our industry is largely systematized around AI at the core, but today we're in this sort of level two. It's an assistant bleeding into potential teammates. And so I just think that framework for us helped our team really anchor around how do we put it into our own language that's not intimidating that we can understand. And so when we think about thought partner, assistant, teammate, system, we could then begin to actually unpack all of the great insights that Brock just shared.
Shaun Tucker: Yeah, I love that. It also just puts you on that roadmap to give you permission to kind of iterate your way there and not feel like you have to take a quantum leap and bite off more than you can chew. A couple questions that have come through, maybe we can do speed round on this. One was the confidentiality of data. So private information, “How do you think about making sure that's secure?” You mentioned that a little bit earlier, Nate, maybe a comment there. And then the other one is for young professionals entering our business, “What level of acumen do they need on AI and what advice would you give them in terms of vocational training?” Those are two totally different questions, but maybe we can speed round and then we'll move to our third segment.
Nate Angelo: Yeah. Why don't I tackle the first one in terms of what we're doing around security and then I'll let, Brock can chime in as well and then answer the young professional. So for us, we were ready to jump into AI and thank goodness we were partnered with a great outside firm that knows our industry in and out and they said, "Hey, slow down a second. Let's evaluate one, where your data lives today, do you really own your data or do your tech partners own your data?" That's simplifying the conversation, if you will. And then it was, “Let's make sure your data is a couple of things. One, it is truly secure and it's in your own environment. Number two, let's make sure that your data is prepared for the SEC requirements around AI application.”
And then number three, the last component for data for us was ensuring that we went to an Enterprise Claude relationship. And it's not to say that Claude is the answer. I think Brock did a great job of outlining the various partners that you can partner with, but that enterprise level puts in place your own sort of sandbox, if you will, that protects your information and data. Where we were actually at greater risk is where we had people exploring with various AI tools, Claude being one of them. And if they were to have potentially entered in actual client data, we could have been exposed. And so we moved quickly.
Not because we wanted to kill the aspiration, but we wanted to put in place guardrails around our data to protect it. So a couple of steps that we went through, and the outside consultant really helped us get that in order in very short order.
Shaun Tucker: That's terrific. I think, again, a systematic approach, very thoughtful. Be mindful of compliance and risk management. At least do no harm is what I'm hearing, but it's also possible to move forward in these ways if you do it thoughtfully and engage outside experts. I think that's clear steer from you. Brock, on young professionals entering the workforce and AI training, what would you recommend?
Brock Sutton: Yeah, so I think, one, this is a way, I think, for younger professionals to potentially differentiate themselves, is to lean into this technology and understand how to use it. A lot of the firms that we're working with, they have this question, which is, "How do I staff this? How do I staff AI?" And I think there's the different kind of approaches to this. One is designate folks on the team that are responsible for it. Typically, what I see is they are a little bit younger, and they have some interest within this space. Another one is kind of the route I think Nate went, which is the external kind of consulting route paired with that internal. And I think we are seeing more and more firms, not just within financial services, but broadly, have this approach. This concept of a forward-deployed engineer, which is basically ... Think of it as like an expert in AI that has done these types of deployments across all companies coming in and implementing.
And then I think the last bucket is to hire people with this skill set. So hire people that have some kind of software engineering background, but are also interested in finance, and try to bring those people into the firm. And I think these are all viable options, really, depending on what you want to do. But I do think for that younger generation, that younger individual, I do think it's something that if they can lean into these tools, have comfort using them, it is really hard to take an entire economy, entire industry, and rebuild it with AI. There's years, there's decades of work that are actually going to go into that. And so if you can come in with that skill set of saying, "Hey, I understand this technology. I understand how it works, and I understand how to build things. And oh, yeah, I'm also interested in financial services, wealth management, retirement," whatever it might be, I think that's a powerful kind of duo.
I think the one risk I would say is you don't want to, as a younger individual, or anyone, to be honest, you don't want to outsource your thinking to AI. And so I think one of the big concerns with this is to say, "Hey, I'm going to have it think through all of these problems, and I'm not going to learn the details." And I think that is a miss. Because, to Nate's point, there's so much value in having an expert in the loop, and then having them double check, validate, especially with where we're at in the evolution of this technology.
So I would say lean into the technology, but at the same time, make sure that you're studying the source material, so that you're not just completely ... don't know how a DCF (discounted cash flow) works or something like that, and you're outsourcing that thought process to the AI completely yourself.
Shaun Tucker: I love that. I love what you just said, Brock. A lot of things. The mega theme here is, where does AI play a role? Where do humans play a role going forward? Nate, you said it. AI isn't just integrating the tech stack, it's really consuming it. But it's also becoming a meaningful part of every aspect of your business. And that gets into our next discussion around strategy and encouragement to begin one and think systematically about the entirety of your business. But Brock, just on your point alone, on the human engagement side, this will be the domain of ideation, innovation. That's going to continue to be the repository, the place where humans are adding the most value. Private data is another. Client service is another, to your point, Nate, really leaning in on that human engagement, the trust, that sort of thing.
So there's a lot here, but let's get into the strategy discussion. And Brock, I'm going to go over to you in terms of ... How should we start to think about constructing a roadmap to get towards that strategy implementation with AI? And then, Nate, we'll come to you with some thoughts about, again, what you're doing in your practice and advice you might give.
Brock Sutton: Yeah, so-
Shaun Tucker: Brock, you want to go first?
Brock Sutton: Sure. So I think one of the things, and this will be helpful for folks, that we're doing when we're consulting with different teams is we are pretty quickly able to evaluate where their firm is at in this kind of AI maturity blueprint. So what I would be doing here during this section is say, "Where are we at as a team? And then what are the one or two or three things that we need to do as a firm to get to that next level?" So first up, everyone starts typically in this kind of ad hoc adoption phase. This is where you have access to some of the AI tools, and you're typically using them in different capacities across the team. Some people might be really bullish. They're using it every day. Other people are not. Level one is where I would say you're not necessarily taking AI seriously as a team or as a firm.
And here, the typical applications are very simple: summarization, content creation, fact retrieval. Those are the types of use cases we typically see in level one. As you get to level two, this is where I would say you start to take AI seriously from a top-down perspective. So you essentially say, "All right." And this is what Nate and his team did, is to say, "Okay, where do we want to focus as a team? What are the three things, the three places that we want to help automate in 2026?" You have to prioritize use cases from a top-down perspective.
The other thing is you have to resource it. And typically, what this means at a minimum is having at least one individual on the team own that specific use case. And then the last thing is you have to evaluate and measure it. So that's level two, and what we see in level two is around 90% of teams are in level two.
As you get to level three, this is where you're taking what you learned in level two, so how to systematically apply AI within your existing business operations, and you're applying that across your entire business. And so this is where your team starts to transition more to that AI manager-type role.
So historically, folks may have been individual contributors, where I own these five processes, but in the future, they're managing AIs that are owning these processes. But they're still on the hook for validation, maintenance, things like that. And that starts to happen at level three. And so you're applying AI across your entire business. You're also becoming much more comfortable implementing it. Everyone's comfortable using it, et cetera. And we see around the remaining 10% of teams that we work with in this level three or scratching at this level three.
Level four. So if you're doing the math at home right, that means we have 0% in level four. And so level four is where you're rethinking your entire process with AI in mind. And so you're saying, "Based on what we're trying to do, how would we rebuild it within AI?" And one of the reasons we just haven't seen firms quite get there is because the technology's moving so incredibly fast. So one of our assumptions is it's going to take time before people get to level four. The technology needs to slow down, because the way that you would build something two years ago is very different than where you'd build today. So what I would be thinking about is, "Where are we at, and then what are the couple things that we need to do to get to the next level?"
Shaun Tucker: Great points, Brock. Nate, can you riff on that? So as you think about your practice, thoughts, comments in terms of where you're at in that continuum and what you would advise.
Nate Angelo: Yeah. I dream about being at level four one day. I think we're somewhere between bridging between level two and level three, and I think this slide that you're looking at right now fits very closely with how we've gone about aligning ourselves to the various tasks that we want to focus on. I think it might be helpful for me just to give the quick blueprint around tactically what we did towards integration, and then I'll summarize it with a couple of outputs. Number one is we formed an AI committee. So we took a business leader from each of our business units, essentially sent them away, and said, "Hey, go build your best AI ideation page. Bring it back to the group." AI committee with our executive leadership team then started to flush through highest value, highest effort, low effort, and began to think about where in which we could start our process.
We then ran that through just a very basic way of the way we think about the business. Is it above the line or below the line? Meaning, is this going to be above-the-line revenue growth advisor team empowerment, or is this below-the-line margin expansion, operating efficiency? And then we started mapping that roadmap for which we could then start moving towards deployment. And so for us, it did start with some very basic tasks, which we've talked about already today around our team: an operating system which you log into every day, and this is really your engine into composition wealth, followed by the basics of onboarding clients, meeting prep, meeting follow-up. And now we're at a place where we're taking tax planning, estate planning, and what we call core financial planning, running a coding system over the top of that to actually pull disparate information into a totally revamped client output that is really a deliverable that is far more dynamic than what we've been able to deliver prior just via a single planning tool or an estate conversation or a CPA conversation. So we're now internally able to marry these worlds through a dynamic output tied to some of the work that we're doing. And so that, I think, correlates a little bit with the various slides that Brock just really hit on in terms of how that's played out for us, both in terms of our integration plan and now where we're at with output.
Shaun Tucker: The journey you've been on, Nate, has been extraordinary, just to see the quantum leap that you've made. I know you felt like you were late to the party, but boy, I don't feel like that, just based on your systematic approach. Very, very impressive. Okay, gentlemen. I know we knew that we were tackling a huge topic here, so we're at time. I'm just going to do a speed round. And just the audience, there are several other questions that we will be sure to answer and send to you, so thanks for the engagement. But first, Nate, to you, then you to Brock. Final words of wisdom on AI, and then we'll wrap.
Nate Angelo: This is just simple philosophy on life. It happens to apply in AI. This idea of being directionally correct, directionally correct and not precisely wrong, has been our MO. We've just anchored to that and recognized, "Hey, if we're moving in this direction, we're going to be okay." Precisely wrong is meaning I'm paralyzed and I'm not taking action. And we just have anchored to that idea, and it's worked for us.
Shaun Tucker: Yeah, it's never too late to get in the river. And you can catch up and innovate, and you'll be in a good spot. Really appreciate the positive encouragement. Brock, over to you.
Brock Sutton: Yeah, I would echo that, and I would say be flexible. We are still in the phase where we are filming radio shows. The change is not going to stop, and how you build these things, how you implement isn't going to stop. So I would say remain flexible. It's really important to apply it, again, because of the economic value that it can provide, but you do need to remain flexible, because this thing is going to continue to evolve and change.
Shaun Tucker: That is terrific. I'm going to just add my piece that, obviously, building an AI strategy is important. You've heard that really in spades today. We're standing by at Capital Group with our practice management team to help you with that, not just where AI plays a role, but where you can optimize your human capital and that important engagement to build trust and credibility with clients. So please do let us know how we can help. Gentlemen, such a great discussion today. I think we covered an enormous waterfront. There's so much more to cover here, and we'll continue this conversation, but great discussion, practical wisdom, application guidance around building an AI strategy, and I know everyone that's on really appreciates it. So great stuff.
Couple quick reminders as I conclude. Number one is please check out Capital Ideas Pro.
We have a number of resources there, an AI page that we refresh all the time with the latest insights and articles. You can bookmark our prompt library. Check it out. It's really helpful in terms of getting started at the application layer for AI. And then more broadly, as the best want to get better, if you're looking to build and lead a high-performing team, check out our digital course. It has a number of really good resources, self-assessments, case studies, tools on what the best are doing. Help unlock that potential of your team and check out that course to do so.
Mark your calendars. Another thing just to let folks know about, our next webinar for Capital Ideas is on July 9th. It features our wealth strategy team. Leslie Geller, Lauren Liebes from that team. We also have a special guest wealth manager Katelyn Wheeler. They're going to be talking all about Trump accounts, as well as saving and planning for kids and grandkids, so look forward to that.
And last, I just want to thank you for your engagement. It's because of your support, our advisors and the teams that are on, that Capital Group was voted number one in thought leadership six times in a row, and also has been named among the top providers for practice management for three years in a row. So on behalf of Nate, Brock, myself, our entire team at Capital, I just want to thank you for your time today, for your engagement, for your support. Have a great rest of your day. And I also just want to wish everyone a reflective and a meaningful Memorial Day. So with that, gentlemen, thank you. Have a great day.
1 hour CE credit for IWI and IAR*
What are your biggest questions about AI?
Bring your AI questions to this exclusive webinar for financial advisors. Get answers, tips and tools to help implement impactful AI strategies in your practice — today and in the future.
Featuring Capital Group AI specialist Brock Sutton along with Nate Angelo, the CEO of a thriving RIA using enterprise-wide AI, our panel is ready to provide answers you can use: From where to start and which tools to use, to integrating technologies across your team.
Event date:May 21, 2026
Shaun Tucker is a director of practice management at Capital Group, creating programs and tools to help intermediaries grow their businesses. He has 26 years of investment industry experience and has been with Capital Group for 25 years. Previously, Shaun led Capital’s sales enablement in North America, and has also held sales leadership roles as a national sales manager for the institutional business and as a division manager in the retirement plan business. Before Capital, Shaun served as a Surface Warfare Officer in the U.S. Navy. He holds a bachelor’s degree in environmental studies from the University of California, Los Angeles. Shaun is based in Los Angeles.
Brock Sutton is Head of Emerging Client Capabilities at Capital Group and has 14 years of investment industry experience (as of 12/31/2025). He holds an MBA from UCLA and a bachelor's degree from Nebraska Wesleyan University.
Nate Angelo is the CEO of Composition Wealth and has 24 years of investment industry experience (as of 12/31/2025). He holds an MBA from Seattle Pacific University and a bachelor's degree from Texas Christian University.