Transcript
Amit
Hello everyone and welcome to yet another episode of Build AI podcast by GI Guy Ventures. Today I have with me Vineet Moan, who is also a part of Guy Ventures and leads many things in Guy Ventures, including international business. And he has been diving deep since last year into the wealth management RIA space and recently spent very nice time on the beaches at Miami meeting a lot of RIAs and wealth advisors.
You know, chilling on the Miami beach. I'm sure they also talked about business a little bit, just a little bit. So, yeah, this is very interesting because I think vertical AI is very important for Guy Ventures. And you know, we have been building a couple of products in this space. While the entire VC community says that focus on one thing, we are madcaps and we feel like diversification is our core competence. I'm half joking there.
But it's interesting to look at AI. It's so early and the technical part of it is changing like every week, every month so much. Like there is Manus yesterday, and then there is another one today. The foundational layer is changing, middleware is changing, application layer, agentic systems, everything is changing. And we still don't know how the broader industry is going to adopt it. There are signs of pilots, POCs. If you divide by large, medium, and small enterprises, the adoption curve is very different. The US is like 10x ahead in terms of adoption as compared to anywhere else.
So I feel like there is so much to be learned. There's so much experimentation has to be done, that it's really good to spend time in a bunch of these industries, understanding what problems can be solved with AI. This is our attempt at Guy Ventures to not only build and sell these solutions in these markets, but also discuss here on Build AI so that other builders can learn from it, investors can learn from it, and hopefully some of the industry practitioners also kind of see it.
With that, I'll just jump in. Basically, we know that the US wealth management and RIA industry is huge, right? I read some crazy number, like some 47 or 48 trillion assets under management. It's unbelievable. The numbers are crazy. And you know, I'm sure you can talk about how many people are employed and those kind of things. But if I was to just look at the question that I want to ask, it's: What is exactly happening in the US RIA industry? People talk about changes are coming, there is growth, but then there are challenges as well. There are margin pressures, there may be talent issues, retirement issues.
I know a few things from there. I spent time at a platform like AngelList called Proplexx, just advising them a couple of years back, and I looked at it from an alternative asset industry and how that is managed. It's slightly different, but I think the licenses and the structure of the industry is pretty much the same. So I know a few things from there and from all the research we have been doing. But you have been very close to it since the last year. You have met with dozens and dozens of RIAs, you have spent time solving the problem now with AI solutions. So can you talk about it a little?
Vineet
No, thanks Amit. It's great to also be on the podcast. Like you said, I think the last three, four months have been very interesting, very revealing. I think like with every other industry and the US economy is the biggest economy, the US financial services industry is the largest financial services industry. But if you look at the very specific wealth management industry, to your point as well, again, it's very vast, it's very deep as well. I've been in banking for a very long time prior to this, but having spent more time in the wealth management industry has helped me kind of peel the onion a little bit as well.
I think several trends have happened in the last few decades, so to speak, but I think right now is a very, very interesting time in the industry, and I'll get to why that is so. Firstly, I think, just in terms of numbers, I think you're not far off in terms of the assets under management. Massive. There are like 30,000 RIAs, 300,000 financial advisors, so you know, big, big numbers. And you've got firms all the way from a Morgan Stanley, which has thousands of advisers, to a sole advisor shop. So the contrasts are quite wide as well.
But some of the interesting trends that are happening is obviously in the US there's a massive generational transfer of wealth firstly. So estimates vary. Some people say 80 trillion, some people say 50, 60 trillion, which again resonates with the number that you said earlier about assets under management. That's how much is moving from the baby boomer generation to Gen X and millennials. So that's a massive shift, and it's not just movement of assets under management, right? It's also movement in terms of investment preferences. Gen X invests, millennials, as many of us know, might like different types of assets, might like to be serviced differently. So that's causing some unease as well in the industry. So it's an opportunity, but also some unease. That's one.
Also linked to that, the fact that if the investors are largely baby boomers, so must be the advisers, right? So the average median age of advisers, again, there's no strict number, but it's probably somewhere in the 50s. Which means that there is a generation shift happening in that space as well. So just to kind of point to a couple of trends there. There is this estimate that with a large number of financial advisers approaching retirement age, or an age where they want to shut shop and move on, there's a big shortage of talent, of people coming in. There's a general sense that there's probably around a shortage of 100,000 advisers in the next 10 years, which is a big gap to fill as well.
And the third thing I would say, I mean there are a few other trends which we can get into as this conversation flows, but the other big thing that we're seeing now in the market is a fair level of consolidation. So there's been around 200, 300 deals that have happened, M&A deals, in 2024 alone, and that trend continues in 2025. This is a big private equity play. So 70-75% of those transactions have been PE-led. In many cases, there's buyouts happening. In many other cases, there's minority investments coming in where smaller firms want some liquidity to be able to grow quicker and better. So it's a very interesting time in what is a large, fragmented industry.
Amit
What about the margins? Like, I hear a lot of margin pressures in the industry.
Vineet
Yes, absolutely. I mean, this is something when I speak to advisers as well and folks in the ecosystem. Obviously, we're talking more about from a fee-based model, right? The RIA is obviously a fiduciary to the client, and the revenue model is largely based on what you charge as a fee on the AUM that you manage for your client. That was over 100 basis points previously. That is starting to shrink now. There's tremendous margin pressure because firstly, at the lower end, there's some level of tech play coming in with robo-advisors and so on and so forth. But then there are also new entrants coming into the market or, with consolidation, there are players coming in who have better efficiency. So there's kind of margin pressure coming in from that standpoint where fees are getting compressed. So if it was 100 basis points earlier, it's probably shrinking towards 70-80 basis points. So that is one issue on the revenue side. But equally, on the cost side as well, right? Because if there's an advisor shortage and you need more people, then you need to start paying more for fewer people that everybody wants to hire. So the revenue and cost challenges are creating additional margin pressure as well. So, yeah, absolutely, that is also a big issue.
Amit
And what are the other challenges that the industry is facing today?
Vineet
Yeah, I think largely I would say the points that I've touched upon are the kind of core challenges. But I think there are preference shifts coming in as well, right? So largely, previously a lot of the investments have been in the public market space. Increasingly, private market investments are also of interest to a lot of the investors, and this could also be because of the newer type of investors who are coming in from a demographic standpoint. So there's a large topic around convergence of public and private markets. So there are new platforms who help organize investments into what the industry calls "alts." So this could be PE fund, a VC fund, private credit, that kind of stuff.
And I think compliance is the other big topic as well. There's a huge focus from a regulatory standpoint to make sure that everyone adheres to the compliance standards when it comes to record-keeping, when it comes to communication with investors. And also our favorite topic, AI. With the use of AI as well, there is an interest in making sure that these technology tools that you're using are compliant as well. So I think that is another interesting topic that everyone kind of brings up. Again, going back to my earlier point, adding to costs as well and leading to further margin pressure, right?
Amit
Very interesting. And you spoke about AI. I wanted to first ask you, just from basic principles, how do the people in the RIA industry see tech in general? And what is the level of tech penetration, openness to adopt new technology?
Vineet
Right, right, right. So I think this is a very interesting question. In the industry, there is this map. It's called the Kitces Map. It's basically—everyone in the space has seen it—and what people say is that every year it becomes denser and denser. For people who might be unaware, the Kitces Map basically lays down all the tech or fintech players in each vertical area, so it could be financial planning, CRM, portfolio management, so on and so forth. Interestingly, I was reading somewhere that in 2018, the Kitces Map had 180 entries and today it has 500 plus entries. So obviously, the Kitces Map is growing because a lot more people are entering the market.
But going back to your earlier question, I think financial tech adoption varies. There are some who are very tech-forward and some who look at it with a level of skepticism. But having said that, there is a level of tech that is definitely required in the industry. What I mean by that depends. If you look at what financial advisers do, right? They provide services like investment advice, financial planning, portfolio management. For a lot of these activities, there is a level of tech that is required. So you have financial planning software, you have portfolio management software, you work with custodians and their software, you have CRM. From that standpoint, everyone ends up having three or four different tech tools.
Now the question is, who is going beyond that and using relatively newer tools like AI, etc., to make their processes even more efficient? That's where I think there is a distinction between the early adopters and perhaps the laggards in some way.
Amit
Right. Okay. The other thing I wanted to ask is, apart from technology in general, what do they think about AI? What is the level of awareness? Do they know, for example, Morgan Stanley gave this AI tool to all its 16,000 advisers? What have you seen in your numerous discussions with the industry?
Vineet
Sure, sure, sure. I mean, again, I think it's a range. I think we've both followed the Morgan Stanley announcement last year with a lot of interest when they launched, I think it's called "Debrief," a meeting notes assistant for all the thousands of advisers. That is a great sign, right? Because if the largest player in the industry is adopting AI, then that is a good signal to the rest of the industry that, hey, this is something of interest. And this is something I'd love to learn from you as well, how other industries are kind of looking at this and whether adoption is coming from the smaller firms or larger firms. But yes, Morgan Stanley was one of the early adopters with the OpenAI collaboration.
But on the other hand, a lot of smaller firms are also utilizing AI to some extent, and it varies. There are some that look at AI with a lot of skepticism because there's a fear that—and fairly rightly so, right?—because you're dealing with sensitive client data here. So you want to make sure that it's kept securely, it's managed securely, it isn't sent—everyone's question is, is it going to be central models? Is it going to be available publicly? So that is a level of skepticism that people have.
But on the other hand, there are people who are looking to kind of try it out, almost in a DIY format even. For example, using ChatGPT or Gemini or one of the publicly available LLMs to generate their newsletters or to sharpen the content that they're creating is something that a few people have been utilizing. And there have been a few other use cases as well. I would say meeting notes is one. I mean, similar to what Morgan Stanley did, but a smaller, longer-tail RIA utilizing an AI meeting assistant is starting to become more and more common. Some of the other use cases, less so, but people are willing to experiment.
But like with any industry, Amit, I would say there are the early adopters and people who are very pro, to others who are kind of slowly testing the waters, and then there are others who are waiting to see where the industry moves towards and then they'll jump in. But one trend is becoming quite clear, just given some of the pressures that the industry is facing, from a margin perspective, the shortage of advisor talent in the workforce. So more and more people are also starting to realize that tech could help solve for some of that. And from that standpoint, people who are more pro-tech and AI will see benefits earlier than others, in my view.
Amit
Yeah. In fact, Morgan Stanley went to OpenAI to get this thing developed, which I'm sure cost them probably hundreds of millions of dollars. And there's a broader industry which is now looking to sort of use some of these tools. And I feel like it's way beyond meeting notes only. Meeting notes, I think, was the first killer application, and there are a whole lot of companies like Jump and Zocks, but I also feel it's a bit crowded as well. But there are so many things. I would love to know more from you, what are those areas where AI can help, essentially?
Vineet
Right, right. So I think, I mean, let's go back to the Kitces Map. I think it's a good landscape to sort of frame our thinking on this. So it's a very—when we say it's a very fragmented industry, there are two elements to it. One is because it's so vast, there's 30,000 RIA firms, all the way from Morgan Stanley to a single advisor shop, so fragmentation in terms of number of firms. But also very fragmented when it comes to tech. So why does the Kitces Map have 500 plus firms? It's insane when you think about it, right? It's because there are very few unified tech players.
Now, like I was saying earlier, if you look at a typical RIA or a financial advisory firm, there are some core elements that they have to do. You have to manage your client relationships. So you need a CRM. If you want to do some level of financial planning for a household that you manage, then as things become more sophisticated, you start using a financial planning software. You move on from a spreadsheet template to a financial planning software. So that's the second one. Then you have your connectivity. Everybody needs to use a custodian, and as you grow bigger, you want to support multiple custodians. And then you also need to do stuff like portfolio management, right? Actually executing the trades, rebalancing trades, generating reports, statements on a periodic basis, and so on and so forth. Which means that a typical RIA firm will at least use three different software tools. So a CRM, a financial planning software, and a portfolio management software. Some end up using more as well.
Now, there's more and more because, you know, another thing is, as you try to overcome margin pressure, the other thing that advisory firms are doing now is expanding the set of services that they offer. So, for example, insurance is one, tax planning is another one, estate planning is another one. And there are very niche software tools coming in to focus just on those areas. So there could be some firms using multiple, six, seven different tools as well.
So what happens here is, what is the source of truth for your customer? There's your customer data in the CRM, the financial planning software, and the portfolio management software. So one is data entry, because they don't all talk to each other. That is one big issue. And secondly, you have your support team trying to make sense of all of this to pre-fill documents, to open new accounts for an existing client or for a prospect, to prepare for an upcoming meeting, to do an annual review, because you need to pull information from all of these different tools in multiple instances.
In a pre-generative AI world, I think it was hard, even with RPA or something like that, right? It's largely rules-based. You can't really cover all these scenarios. So you had a large level of manual workflow just to kind of work across all these different systems. With AI, you can start solving for some of that.
Amit
Before you go ahead, I just wanted to ask you a quick question, sorry to interrupt. Basically, when I worked for Prove, which is a large company in the identity authentication space in the US, basically we used to use more than like 75 different SaaS tools. And it's very typical of US companies, small or large, tech or non-tech, that you would rather use a specialized tool for each and everything. At least this was very true pre-AI. What I am observing now is that—and I don't know if it is true or not—but I feel like going forward, there may be that one-stop solution kind of a thing which may emerge with AI. Because if you are able to stitch the system of record and these individual applications, and also probably do the work of half of these applications, but also as you said, there were some things which were never automated and were manual in the SaaS world. Maybe those can also be automated, and you can basically have an AI agentic system helping you throughout the journey.
Vineet
Yeah, absolutely. I think that possibility is now becoming more and more real. I think, you know, why? You could still have multiple systems of record, or you could have an overlay system of record generated by AI, managed by AI. You can stitch together—you know what I mean? Think about a prospect that you onboard. You have a conversation with the prospect. AI can listen to that conversation, pull all the information, and then you want to start creating a plan for the prospect and say, "Hey, this is why you should work with me." The AI could use information from the conversation and documents that the prospect has sent to help you build that financial plan. And if that conversation goes well, the next step is onboarding, where again the information that's been previously captured can be used to start pre-filling some of the account opening forms, preparing those client service agreements, and so on and so forth.
So some of those end-to-end flows can really be managed by AI, which wasn't possible previously because you had multiple systems and you had different people looking at different aspects of the workflow. So absolutely, to your point, I think it's very much a possibility. I've seen some firms have tried to do it, to try and become the one-stop shop. I mean, Orion's an example, Investnet is an example, but they again don't cover everything either. But I think with AI, that can change for sure.
Amit
And given that you are developing FastTracker AI, having it under Guy Ventures and have been very focused since last year on that. Can you tell us a little bit about what you are focusing on? What are you building? How will it solve problems in the industry?
Vineet
Yeah, no, absolutely. I think FastTracker AI, the vision really is to become an AI-powered partner for advisers. Like I said, there's a variety of workflows that are largely manual today. Some examples would be bringing together material that you already have about a client ahead of a meeting to plan for a meeting, to type up information during a meeting to make sure that you don't miss anything, to take information out of a bunch of documents that you have from the client or prospect so that you can actually start and analyze it, to pre-fill a lot of forms before you send to a client for their signatures. These are all activities that no adviser or support team likes to do, but they have to do. I would say it's just grunt work. With AI, like we said, we can start automating a lot of these things and stitch together complimentary workflows. And that is really FastTracker's focus as well. The idea is to automate grunt work that people don't want to do and shouldn't be doing in an age of AI, and free up time that they can then use to actually spend with their clients.
Even for support staff, right? I mean, if you have an associate adviser today, largely their time is spent in admin work like getting information from prospects, typing up notes, pre-filling documents. In an AI world, they can help the adviser prospect better, plan for meetings better. So time that frees up from admin work can be used for more productive work that people would prefer doing as well.
I'm also very keen to kind of hear your thoughts on this, right? Because you're looking across industries. What are some of the trends that you're seeing in terms of adoption from a markets, from a products standpoint?
Amit
Yeah, I think one major observation is that the US is at least 10x ahead in terms of general adoption of AI as compared to any other market, be it Southeast Asia or whatever. Except, of course, there are pockets and companies which may be doing work in other regions as well. The other thing that we observed is that the large enterprises in the US are still doing POCs and pilots. So there's no very mass-scale adoption, except maybe in the wealth management industry where Morgan Stanley has actually given out their AI tool to all the 16,000 advisers and they are using it and they are collecting feedback and so on and so forth.
The other thing which I like about this particular market, about FastTracker AI focusing on this space, is that this is the SMB market, as they call it in the US—small and medium businesses. Small and medium businesses are the biggest surprise from an AI adoption perspective. Because one, they do not have a very huge number of people who have to take a decision. In most cases, you are directly talking to the founder or the owner of the business. Secondly, the manual labor is very costly in the US. They are facing talent cost-related issues. So for them, adopting and replacing human labor makes a lot of sense, especially as you said, for non-strategic, non-core activities.
And then the other thing I feel about SMBs which is very strong is that you don't have very crazy regulatory compliance, IT policies, data security—those things which are sometimes we start thinking of problems or creating problems which don't exist and overthink that. So I think that happens less in that SMB segment. So because of that, I feel the adoption is happening first there at a massive scale. For example, I was talking to one of the RIAs which manages about $25 million in AUM, and even they are already using AI tools. So compared to any other part of the world, I think this is very different in the US market.
Apart from that, I feel like from an AI technology perspective, things are changing—I would like to say monthly, but it's almost like bi-monthly, like 15 days, 20 days something new comes up. And because of this, what will happen is that it's very unsettling. So, for example, in one of the products that we are building, my co-founder and CTO Kushel, he had developed—he was using LangChain as the agentic platform for middleware. And then he realized that there may be a better way and to avoid disruption due to changes in the APIs or something or the other that keeps happening. So he built his own. We have our own middleware now, agent system of our own. But now since you have your own, you have to make sure that it is best-in-class and it is up to date and it's ahead of the curve. So now you have to do a lot of this horizontal AI work as well, apart from the vertical AI products. So those are the things I feel are very interesting.
I feel in the language that is used in the RIA industry, the wirehouses—Morgan Stanley has obviously taken a leap of faith and already implemented something. I think in regional firms, it's mostly meeting notes. I feel broadly, meeting notes is very popular. I think independent firms are very interesting, they are actually trying out a lot of different things. So pretty good.
Vineet
No, I think bang on. Right, what you said earlier about smaller firms being quicker with decision-making and the ability to try things out is absolutely true, and it's coming through in the conversations that I'm having with many of these firms and advisers as well. The larger firms, it's harder, bureaucracy, processes are longer, too.
You made an interesting comment there about horizontal AI, and obviously we're seeing the developments in that space, but vertical AI is equally interesting. And I know you've written about it as well. How is that shaping up? And if you were to, obviously we talk about this a lot, but from your vantage point, something like FastTracker AI, how can we win, right, in this particular AI space?
Amit
See, I think, based on our last 12 months of developing vertical AI solutions, or rather I would say, okay, the first few months we were just experimenting with an assistant here or assistant doing that, but last 6 months we have been very focused on these vertical AI solutions at Guy Ventures.
We are, in a way, I mean we have a lot of respect for what YC has done. It's just that our model is like diametrically opposite to YC. YC is saying, "I will get a 25-year-old engineer building a company, I'll back them, and they'll go and disrupt an industry from outside." We are saying that we will go and partner with somebody who has worked in the industry for God knows how many years, and they know the workflows, they know the processes, they know the systems. But they are entrepreneurial, they are already an entrepreneur, they already wanted to do something. And we'll partner with them and provide them the startup building system, as well as the AI tools and technology, and help them build the technology team and so on, and provide them the necessary resources and capital so that they can actually go and automate the industry.
And from my investing experience, I have realized that I have more successes doing these kind of investments than the YC type of investments. So there's a little bit of personal bias and learning as well. And so I, therefore, we are a big believer in vertical AI. So we are mapping out—and we have done a lot of that. Since my Medi days, we have been doing mapping out of various segments and subsegments in financial services. And now, for the last six, seven months, we have been mapping out different segments and looking at what can happen. But then for us, it takes a lot of time because we study, we talk to people, we have to find the right entrepreneurs to be able to build.
I think in FastTracker, it's fantastic how we are able to collaborate. You have 14 years experience at banks like HSBC in the US and Europe, and now you are able to sort of talk to the wealth management industry and understand what is happening, understand their language, understand the problem statements. And I think with the design partners and the customers, it's happening even more in FastTracker.
So that's the kind of vertical AI. If I was to summarize in two lines, one thing is I think in our mind it is very clear that an industry-first, problem-first approach is better than an AI-first approach or a tech-first approach in general. We have seen this movie before in fintech and blockchain and so on, where once you have a hammer, everything looks like a nail. May not be the right way. Maybe it's just, spend more time. It may be boring. It may be slow. But it's always good to first understand where the problem is and what is a painkiller, what is a vitamin, and solve the pain. Build the painkillers for most important problems. So that's the approach we are taking.
I think vertical AI is interesting from that perspective. It's very different. It's a B2B proposition. So it's not going to be as glamorous and tens of millions of dollars raised and all of those things. It will be more organic and slow. And then once you start putting the numbers, that's where the exponential growth starts happening, right?
Vineet
Right. No, I think, I think, like we jokingly say with the focus on experience, I think a 42-year-old is now the new 24-year-old in the age of AI, right? Which is quite true. But that's very interesting. I guess maybe another question was, obviously the tech side of AI is evolving rapidly. So does this become the moat in AI? Then the domain expertise, what really is the moat in this world?
Amit
Yeah. Today, if you see the whole vibe coding thing, if there is an engineer or not even an engineer, now there's a product manager which uses Lovable and can actually create a product. In this day and age, of course, those two products that these guys are creating in 5 minutes and 5 hours or whatever, they're not ready to be given to the customers or clients cannot use them today in most cases. But I'm very sure that slowly and gradually, even those edge cases and issues will get resolved, and we will be able to create at least good MVPs using AI.
So in this kind of an era where it looks like very soon there will be no technology moats—like anybody can take a piece of software or AI, feed it into Cursor or ChatGPT, and say, "Can you create me a replica of this? Just copy-paste this," and it can be done. And then who knows, you will be able to launch a thousand such companies as an individual in this day and age.
I think I feel like we are going a little old-fashioned in terms of what will be more valuable. I think the relationships will be very valuable. Like, if you have very strong relationships with banks and FIs and whoever your customer is or your customer segment is, you have a good understanding of the problem and you know what kind of solution.
For example, I was talking to the CFO of a company yesterday, and he's planning to build an AI tool for FP&A and accounting processes. And he mentioned something very interesting. He said the first product should literally be a plug-in into the Excel sheet, because it's almost impossible to get CPA firms and CA firms to move away from Excel and the accounting software. So you said two things: you can't change on day zero is the system of record, as well as where they are actually developing these models. You don't want to change that on day zero. First, you develop a way, like an Excel plug-in, to pull data from the accounting software or the system of record and build models in Excel sheet. But then from there, maybe 2 years down the line, one year down the line, they feel very comfortable. They'll say, "Okay, what's next? What is the new user interface for this?"
So I think that evolution will definitely value relationships and will value an industry-first, problem-first approach.
Vineet
Got it. Makes total sense.
Amit
I think technology will be easily available to everyone, and that will be level. I think it's what else that you bring to the table that could become the moat.
Vineet
Yeah.
Amit
And not to say that AI expertise is not important. That's why we have a very, very strong technical CTO, Kushel, who is building state-of-the-art technology, document processing engine, whether it is AI agentic systems that we had built, like 23 agents with a supervisor, because now the bar is very high in terms of hallucination and experience and issues that may crop up. How do you proactively solve them? Also, a very interesting thing that we found is that if you build an AI system which is solving a problem today, if you leave it as it is for 2 months and don't change anything, issues start cropping up on their own. So there's a whole lot of evals and testing and continuously evolving the models that has to be done.
Vineet
Right, right, right. Interesting times ahead.
Amit
Yeah.
Vineet
No, this was a fun conversation. Thank you so much, Vineet. And you know, hope you enjoyed your time at Future Proof. It seemed like a pretty good conference. And I think I'm looking forward to how FastTracker AI engages with the broader RIA community in the US. And would love to have some of your RIA friends on this podcast next time.
Vineet
Yes. Yeah, no, absolutely. I think, no, lovely. It was a lovely conference. And I think a lot of the things that we're talking about is resonating with folks because these are real problem statements for them. And yeah, I'm sure we can entice some of them to join us on future conversations. But enjoyed our conversation today as well, Amit. Thank you.
Amit
Thank you. Thanks, Vineet.
End of episode · Ep #03