Money Matters Episode 331- The Invisible Partner: How AI Gives Advisors Their Time Back W/ Arnulf Hsu

Money Matters Podcast 331

Christopher Hensley RICP, CES: [00:00:00] 500 hours. That's how much time one advisor saved just by changing the way they run th their meetings. No extra staff, no more clicking through screens. Just talking. Today I'm sitting down with Arnold Schu, the founder of G Reminders, who built a voice aware AI system that does what every advisor wishes their tech stack could do. Listen, remember, and act. His system already in use by top performing advisors is freeing up entire work weeks, cutting admin time in half and making CRMs feel like they actually serve you. But this isn't just about automation. It's about getting your attention back so you can show up fully present in every client conversation.

This interview is part of my digital Kaizen for Financial Advisor series, where we explore the small shifts that lead to big wins if we've ever wandered. What's possible when your tools finally work with you, not against you. This is your blueprint. Arnold, thank you so much [00:01:00] for joining us today. Absolutely. Absolutely. And you know, we, we, we did an, uh, article on this, um, for the book that I have coming out, the Digital Kaizen for Financial Advisors. But this was a, uh, a deep dive on this topic of, of G reminders and. Um, and so I'm gonna ask you some questions based on that, but I encourage people to go out in LinkedIn and find that article.

We got a lot of eyeballs on that when we, when we put it out there. Um, and for, for other advisors who are, you know, considering what, what do I use for my tech stack? There's a lot of choices there. Uh, one of the things that you mentioned in that article that, that really stood out to me was that you, you talked about how AI can save advisors up to 500 hours a year. That's a staggering claim. Can you walk us through what that looks like in an advisor's real world workflow? From that moment, a meeting is booked right up to the follow up.

Arnulf Hsu: Absolutely. Um, by [00:02:00] the way, I, before I get into that, I, I love the, i, I love the sort of digital kaizen term, you know, Kaizen is, is sort of continuous improvement and, and as you improve different aspects, you really, you really get very, very meaningful. Um, output and improvements from your systems and processes and so forth.

So let me talk about sort of, you know, what we do and, and generally we, we look at everything through the lens of a meeting life cycle, right? And so if you think about wealth management and what, what financial advisors do, do they effectively, you know, spend time with their clients, um, getting them to understand, um, how their portfolio is performing and, and ultimately, you know, leading a better, um, sort of financial life.

And so when, when you look at that sort of life cycle, they meet with their clients typically maybe twice, twice a year, once a year, or maybe once a quarter. Really kind of depends, um, on what their cadence is. We deeply integrate with the systems that they use every day. So typically CRMs and, and, and other systems that they interface with [00:03:00] Outlook and, and web meeting tools and so on and so forth.

But we understand the cadence that they have with their clients, right? So it starts with that. Um, what, what's the frequency you're meeting? We look into those systems and we automate the process of actually getting something on the calendar. Um, so it sort of starts with sort of the scheduling aspects of things.

Um, making sure that we understand the availability of the advisor, the availability of the client, and effectively facilitating that, um, and giving some self-service tools to the client so they can, they can, they can book a meeting. Once a meeting is booked, you now have different workflows that might get kicked off within a, within a CRM or with a, with other, um, sort of process tools, um, that might kick off certain actions and different things that need to happen.

Um, so that's sort of effectively all automated. We also make sure that the client actually shows up to the meeting. We send them email, SMS different types of sort of communication channels to ensure that they show up, um, furthermore before the meeting, right? So typically the [00:04:00] advisor needs to get better prepared for the client.

Um. Do a, a sort of recap of, of what has transpired since the last time they talked could be in the last six months, last 12 months, all the things that have happened could be many variety of things in including portfolio performance or other things that have gone on in the market. So we put together a very nice pre-meeting brief that is effectively custom right per client using generative ai.

Um, so that the advisor is doing less. Easy work before the meeting. Essentially aggregating all this stuff across all these different systems that they're using, right? So advisors to use 5, 6, 7 different systems. Could be custodial platforms, could be financial planning tools, you know, it could be notes, could be documents, and so on and so forth.

Aggregating all that stuff together, providing a better pre-meeting brief, again, leveraging AI and automations. And then once you're in the meeting, actually having the meeting just like we are today. Um, we have a notetaker product that [00:05:00] essentially joins the meeting. It's effectively a bot, um, that listens to the meeting, transcribed, and then pulls out summaries action items.

So after the meeting, so, so the advisor is completely present right when, when they're, when they're engaged with the client. Um, and then the system is effectively taking notes. So before maybe you had a paraplanner or something participate, they are taking notes. You don't need to do that anymore as effectively AI is doing that on your behalf creates a very nice meeting summary.

Um, financial notes, uh, you can lay out sections, customize, et cetera, et cetera. Um, it recommends action items for different people. So you talked about Chris, you know, go do, go do X, you know, to, you know, take your RMD, et cetera, et cetera. Um, and then again, pushing all that stuff back into your system of record so that it's effectively all there.

And then lastly, we also pre-draft a follow-up email for the advisor [00:06:00] in their inbox that's effectively sitting in the draft folder as a follow-up to the client. Say, here's what we talked about and here are the things that we sort of actioned on. And the advisor obviously can review that, make any changes, and then quickly hit send.

So hopefully, you know, with all of those different aspects, you can kind of see. You know, some of the impact on, you know, time back to the advisor.

Christopher Hensley RICP, CES: I love it. I love it. Everything that you kind of touched upon here, you know, as an advisor, the idea of it being calendar centric or meeting centric. Uh, you know, I think I learned a long time ago in pro productivity philosophy like. Stick to one calendar. Right. Don't have multiple calendars. So having something come out of that one calendar to kind of kick it all off is, is fantastic.

The other thing is, I shared with you before we started this, right? I just got back from taking my mom to the doctor's office. This is relevant. I'm getting there. So, so the idea. Of being able to, you know, sit in somebody's office and [00:07:00] have you ever been at the doctor's office and you, you just expect them to have looked at your medical records before you get there, and then you get this blank stare and, you know, they didn't look at it. But from what you're describing with G reminders, it's, it's literally. us these pre-meeting notes where it's doing a nice summary of those conversations so we're not having to, to reinvent the wheel before we sit down, putting us in a really good spot before meeting with the clients. I love that.

I love that. Now I'm gonna, uh, pivot and talk about something that you wrote about in this article, or that you talked about in this article, and it's the invisible partner concept. Use this term, the invisible partner to describe your ai. I love that framing. Uh, and it feels less like replacing advisors and more like empowering them.

How do you see that mindset shift changing the way advisors approach their tech stack?

Arnulf Hsu: Yeah, I, we definitely think of AI as a partner, certainly not as a replacement. Right. And if you are. Um, if [00:08:00] you've accumulated some wealth, you know, over time, uh, you are, you are unlikely to give those decisions to a robot or AO advisor, as is commonly the case for sort of lower net worth, um, folks. And so to that extent, you want to empower the advisor ultimately.

That is why there's a, there's a, there's a service engagement in the first place. Otherwise, what is the form? So AI is really there to sort of help, um, the advisor. Um, you know, there's, there's a lot of places where AI can help and assist, right? So we think about AI and automations as not just sort of AI on its own.

Um, but there's a lot of places where, where AI can help. And so, you know, it, it really starts with. Um, minimal change management, right? So change management is a very difficult thing in general. If you ask people to change how they behave, that that is just sort of tough to do. And so for the most part, we try to minimize change [00:09:00] management in, in the sense that how do you work today?

Okay? Let the AI and automation sort of do some of the things that you're already doing manually without changing anything. Just don't do them. Let AI and automation take some of that off of your plate. Um, but you're still engaging with the client. You're still meeting a client, you're still sending 'em notes and, and emails and so on and so forth.

But again, let AI help you sort of make you sort of, you know, superhuman as as, as we like to say. And so for the most part, we've kind of focused a lot on kind of back office stuff. So, um, you know, low impact from a client standpoint, but high impact from a, from a back office efficiency standpoint. But we're starting to see some of that stuff creep also into the front office it with, with, with some of the email drafts and, and, and, and so forth.

And, and I think you'll, you'll start to see some of that come through in, in various, you know, marketing capabilities, um, down the road. Um, but we've fo focused, I would say mostly on, on sort of back office, um, you know, high [00:10:00] impact again for, for, for, you know, back office work and, and, and low impact on the client side.

Christopher Hensley RICP, CES: that. I love that. The idea of change, manage it, but minimizing that, right. That's. With anything people, you know, that if you ask 'em to do something really, really big, it's less likely it's gonna get done than things that are frictionless or are less friction. So I like the idea of that being integrated into this whole process here. Alright, I've got one about compliance here. And so, you know, what I'm finding out, um, as I'm writing this book, uh, digital Kaizen for Financial Advisors is one of the things that's on the top of everybody's mind is compliance. For me. I don't even like. Talking about it. 'cause I enjoy all this other stuff, right? but this is a conversation that has to be had, so you know what I call it. I want to get behind beyond the cover your own ass thing. Right. So, you know, we saw that attorney that went out and wrote the brief using Che GPT, and we've been kind of playing. You know, people in the healthcare profession, finance, we've been playing catch up [00:11:00] from then, so we kind of gotta acknowledge it and speak to it.

But when it comes to our industry, uh, and compliance in the article, you said the biggest hurdle isn't the tech itself, it's compliance. What specific concerns do compliance teams raise and how did you engineer g reminders to address those without slowing innovation?

Arnulf Hsu: Clients always is part of the conversation. Um, and so we, we, we understand that, um, especially in the financial services space. Um, compliance is a big deal, so we spend a lot of time effectively leaning in to compliance and governance so that the folks that are effectively in charge of tho of that discipline are comfortable with those systems.

And so that includes understanding our systems and processes internally. Um, we have third party audits, for [00:12:00] example. So we, we, we, we do, uh, soc two, uh, compliance effectively. Um, so, so that it's effectively what we're doing is, you know, third party audited. Um, we talk a lot about data privacy. We talk a lot about AI not learning.

On, uh, data that is, that is provided to, you know, the new LLM and, and new generative, um, sort of AI technologies. Um, so we spend a lot of time sort of, you know, in this, in this data security privacy space, again, to get folks comfortable. And the other thing that overlies, I, I would say, you know, all of this.

Is again, that AI is an assistant, is a tool, but ultimately the human is still in the loop, right? And the human is ultimately making the decisions. And there's always that confirmation back to say, is this what you want to do? Confirm, right? And it's not just AI sort of, you know, going rogue. And, and, and that's, and that's something that I think.

You [00:13:00] know, when, when you're working with ai, make sure that you sort of, you know, understand the, you know, who, who's, who has access to the data and what is happening with the data, and what actions can be done on top of that data set.

Christopher Hensley RICP, CES: that. Great

Arnulf Hsu: I.

Christopher Hensley RICP, CES: there. Uh, you set me up for the next one here that I'm gonna ask you about human-centered and efficiency. But before we get there, uh, you know, some of the things you just shared with us is leaning into governance, uh, you know, for the people that, that are, uh, doing compliance and making sure that we're. Answering those questions, it's so important that the technology that we're working with, that you guys are on the same page. You know, I'm with a, uh, a regional broker dealer, but even if you're at an RIA, these are the same issues, right? Uh, the idea of having SOC two compliance, so if you're an advisor and you go get very excited about some technology and it's just not able to check the box on the things that are required for data privacy. It's not worth going down that road. So it's good to know that you're working with a, a [00:14:00] partner that's, that's aware of these, these, um, guardrails that we have on it. And then, you know, you, you also mentioned not ai, not going rogue, right? I even picked up when you, uh, prepared the, the emails for us. It didn't just go out there, you put it in a draft folder.

So there's that. Human touch to it before it gets out in the world, which, you know, as, as somebody who's sending an email out, we should be doing that anyway. Right. So, so, uh, super important. So let's talk about presence. You said AI freeze advisors to actually be present with their clients rather than worrying about notes or follow up.

How does that translate into better outcomes for both sides of the relationship?

Arnulf Hsu: I think it, you know, uh, wealth management is very, is very, it is a, it's a high touch. Um, bespoke business, right? Um, that is why you're engaging a financial advisor for them to understand you, understand [00:15:00] your goals, ultimately, um, to lead a better financial life. And so as a result, you really wanna understand your clients.

You really understand what makes them tick, what they care about. Um, you know, what are, what are, what are, what are the things that, that they're trying to achieve? And so in order to do that, you really need to be present. You, you, you shouldn't be, you know, writing down notes and, you know, the human, in fact, cannot, cannot write and cannot read.

And talk or listen at the same time. Like it, it's actually not possible in our minds. And so you really wanna be present. And so to the extent that others, you know, in the past, like I said, maybe you had a paraplanner or somebody else present that was sort of, you know, doing a lot of the sort of administrative work.

Now effectively, you know, AI can do some of that stuff. And those para planters, by the way, can now. Uh, uh, spend time on higher value activities rather than some of the sort of rote admin work. And so, again, you know, we think that AI and [00:16:00] automations allows the, uh, uh, allows the, uh, the, the, the advisor effectively to be really, to be present with their client, which is ultimately really what they should be doing and what they're good at.

And I think what they want to be doing.

Christopher Hensley RICP, CES: I love it. I love it. Yeah. I, my experience when the note taker ais came out was that, uh, I don't know if, because I'm an obsessive note taker. If I'm really being honest with myself and I, how many times do I actually go back to my notes and then it's chicken scratch? I can't read read what I actually wrote, but when I started using the AI note takers it, it really does a really good job of capturing the things that are important and even the things you didn't think were important, but you may need to go back.

Future time to see. So, so I really love that aspect of it. Now I'm gonna keep moving here 'cause we've got about, uh, eight minutes here before we hit the end of the show here. Uh, but there's something that you have in g reminders that's called the ask and do Anything feature. That's a powerful feature. But you're also clear about [00:17:00] how important context is to its success. How do you ensure that AI's actions are relevant, timely, and safe inside of an advisor's ecosystem?

Arnulf Hsu: Yeah, good question. Um, so context is everything to a large language model, right? Generative app. It really is everything. And so. Um, the way that we think about context, we think about it in terms of a client record or a household, right? And so we don't actually, today, we don't, today we don't have a general AI assistant, right?

So we have effectively an agent or an assistant that sits on top of the household or the client record. And as a result, it's effectively guardrail. Just on that. Um, and so the client id, the email addresses, things like that, things that are sort of effectively, you know, primary IDs or primary keys help, um, with the guardrails as far as what data it has access to and the data that is effectively ultimately [00:18:00] powering the responses.

And so. You go effectively, you navigate to a household and effect. You can ask questions against the entire data set that underlies that particular household. It could be based on past meeting notes. Again, it could be on portfolio performance, it could be on a variety of separate, different topics. But again, guard rate, guard railed on top of, um, effectively the client record.

And, and that really, you know, helps significantly, um. It, you know, the, the most difficult thing is building this sort of generalized AI assistant that is fully aware and has federation to, you know, all these different systems. And that's kind of where, you know, you start to see hallucinations or, or, or, or areas of, of, of sort of going road.

So that's kind of how we've, um, guard railed it today. Um, and again, we'll, we will, you know, make, uh, evolution over time here.

Christopher Hensley RICP, CES: love it. I, you know, when it comes to that, the idea of less is more sometimes and giving it very [00:19:00] good context to where it's gonna go, uh, by, you know, limiting it by household. When you're asking a question, it makes it relevant and it's not going all over the place to track the information down. I love that.

I love that idea. Um. That idea of being able to, to answer from that, and then the idea of just the data set that you're basing it on, um, you know, we are. early on these note taking ais, and I know there's a, a whole bunch of other stuff that's rolling out, but if we go deep just on the note taking stuff, if we project ourself out into the future, five, 10 years from now, we've got all of this data rich.

Uh, recorded information in our client, uh, folders. no way we can't be more of an informed advisor with that much information, but the idea of having those guardrails, I think is, is huge there. We got just about five minutes or so. One of the things that you mentioned. And the article is a shift from reactive AI to proactive [00:20:00] moving from pulling data to to having AI push next best actions. What does that look like practically and how do you avoid overloading advisors with more digital noise? It's.

Arnulf Hsu: So this is, um, this is an area that I'm super excited about. Um, it's definitely, um, an evolving. Area, and we're still doing lots of sort of, you know, active development, um, um, with regard to this, but pulling out market data, combining effectively market data and private data, right? So data around the client around their goals, um, around things that are happening, um, around their financial life and bubbling that up.

Through notifications, through dashboards and different things of that nature, kind of really changes, um, you know, changes the way that, uh, clients, or, or I should say, advisors [00:21:00] have the ability to essentially identify areas that could be of interest or, or, you know, could be, um, issues and bubbling those up and having relevant, timely conversations.

Um, now the challenge here is making sure that we're not fire hosing a data set to the advisor so that, so that it's sort of piecemealed and at the right time. Right? Um, and so there's a lot of work that's sort of happening in this particular area, but it really kind of changes this whole, um, notion of like, you know, clients asking for stuff versus being proactive and sort of, you know, pushing data out.

That makes sense. And so again, it's some sort of evolving area. You know, there's a tremendous amount of work here. We spend a lot of time sort of in what we call plumbing, like it's less important what model, like what, what generative AI models are being used. It's more important that the data, that you have the [00:22:00] right plumbing in place so you can pull the right data elements so that you can ultimately drive the insights.

You know, for the, um, for, for the advisor and ultimately to the client. Um, so we have different types of dashboards and so forth where we're surfacing certain things. And again, we, you know, we continue to obviously work in this area. It's all very nascent, I would say.

Christopher Hensley RICP, CES: I love it. I love it. Great. A uh, answer there, you know, the idea. I, I get digital overwhelm here and the idea of things getting on my radar bubbling up to me, these notifications, uh, having a, a dashboard with the things that are most important that I need to, to be aware of, that's important to me. Uh, 'cause it, it limits a lot of the noise that's out there.

Sometimes too much information is bad, right? Uh, you talked about plumbing and making sure sometimes it's not, uh, the size of the LLM model. I've seen, I've seen a lot of people put small. LLMs on NAS systems and they're, they're a lot smaller, but they can get a lot done. So it's not always the the size.

Right. Uh, the plumbing though, just making sure that you have that [00:23:00] data, the right stuff that it's pulling from is, is huge here. Alright, I think we've got time for one more question here. Uh, a big part of digital Kaizen is about starting small. You recommend note takers as an entry point. Why is that a good first loop and what kind of results do advisors typically see early on?

Arnulf Hsu: So I talked earlier about change management, right? And so we think change management are really, really difficult, especially when it comes to business process and so forth. And so minimize the change management. Um, so the reason. So a couple things. One is, so when I, when I talk about note takers, I also include like pre-meeting briefs, for example, right?

Because you're already kind of doing, you're probably doing some form of preparation for the client. Let the system do that for you. If even if it gets it to the 80% mark, that is pretty good and it's literally delivering to you in an email or in a Word document or A PDF that you can effectively then go and modify.

So you're already doing [00:24:00] effectively pre-meeting prep. Let the system do 80, 90% for you. You're gonna be better off. Then note taker, you still have the meeting, you still have your in-person meeting, you still have your Zoom meeting, your Microsoft Teams meeting. Doesn't matter. Let the bot join. It's just another participant.

And look at the output. The summaries are really, really excellent. You can customize those. You are already doing summaries. Let the system do it for you. Let it re recommend action items. And the biggest value is not just that it takes the summaries and, and, and the action items, and, and recommends opportunities and workflow, but it actually puts it in the right place, right.

Like getting that from one system to another is, is work like, let the system do that for you. You are already doing it, but now the system will just automate that. Right. And, and, and. It's stuff that nobody wants to do. Like nobody became a financial advisor to do a bunch of this rote admin stuff. They just did it.

Um, so like their [00:25:00] life literally becomes simpler and the change management is just do less of that stuff and let the system do it for you. And that's why we say start with that because it's almost instant gratification.

Christopher Hensley RICP, CES: I love it. I love it. And that, you know, some of this can be dangerous. You as an advisor, we need to be building relationships, making contact with our clients. Uh, if you're at all, uh, like me, I, I enjoy. Building out Zapier integrations. Well, I shouldn't be doing that. That is a rabbit hole. A time suck. And half the time, you know, three hours later you're like, okay, what the heck am I doing?

So if there's something that where, where these things are already integrated, you mentioned you should already be doing some kind of briefing, but that takes time. And if this is gonna show up with something that's already. 99% there and we're having to go back and do a few tweaks, few edits. You're, you're getting good information there.

Alright, we are right here. I [00:26:00] actually got one more question for you. Alright. This is the last one here. Uh, imagine we are five years out here. Where do you think this is all headed? What's your vision for what an advisor's day will look like in five years? And what role does this voice first AI play in the future?

Arnulf Hsu: So, so the, you know, voice, voice modality is, um, you know, interacting, uh, effectively with your systems and voice is, is sort of, um, underrepresented in general. Um, but I, but I think it's, it's, uh, highly useful, um, uh, sort of, uh, you know, communication channel, um, over, you know, interfacing through a user interface or, or, or typing and so forth.

So I, I think, I think voice will be, will be a big part of it. Um, I think you're going to start to see, well, you're already starting to see, but you're, you're gonna see more and more processes effectively, AI fi or, or, or automated. Um, you know, a lot of the things [00:27:00] that are holding, uh, things back really at the end of the day, we talked about it is.

Access to data and the plumbing that underlies all this stuff. So getting federated to the different systems that, that, that you're using, making sure that that, that sort of, you know, AI or automation fabric is actually connected to those systems in, in a, in a salient way. Um, you know, those are some of the things that hold it back and, and, and, you know, that is, that is effectively all being built out.

So I think you're gonna see a lot more efficiency gains, you know, significantly more efficiency gains. Um, I, I don't actually believe that that's actually all that impactful from like a, you know, labor force standpoint. I think certain areas, you know, may be impacted, um, sort of outside of financial services, but I think a lot of the folks that are doing certain things today can focus on higher value activities.

Ultimately delivering a better client experience.

Christopher Hensley RICP, CES: it. I love it. Arnold, thank you so much for joining us today. We are right here at the end of the show for listeners who'd like to find out [00:28:00] more advisors who'd like to dive into G Reminders, where can they find out more information?

Arnulf Hsu: So go to our website, g reminders.com, and you can also hit me up on LinkedIn.

Christopher Hensley RICP, CES: Perfect. And if you're driving right now, just know that's all gonna be in the podcast notes there. Arnold, thank you so much for being on the show today. Have a good rest of the day.