The Modern RIA: Leveraging AI for Growth and Efficiency

Discover how Helium Advisors partnered with Invent and Finmate to modernize its operations and deliver greater value to clients. In this session, the leaders of Invent, Helium Advisors, and Finmate share lessons on how thoughtful adoption of AI can help RIAs scale, improve productivity, and stay competitive in a rapidly evolving market.

Transcription:
Transcripts are generated using a combination of speech recognition software and human transcribers, and may contain errors. Please check the corresponding audio for the authoritative record.

Oleg Tishkevich (00:08):
So today we're going to have some really fun conversations about AI and RIA. I'm going to start with a little bit of intro. My name is Oleg Tishkevich, I'm the CEO of Invent. And one of the things that I want to call out right off the bat: our company does data. And fun fact—anybody knows what zettabytes are? You guys know what zettabytes are? Anybody? No? It's a billion terabytes. So one zettabyte is a billion terabytes. You guys familiar with terabytes roughly? Yeah. It's a billion of those guys, which is the amount of data that is produced just by our industry. Can you imagine that? The amount of data being generated and analyzed is just mind-boggling. I think they have to keep coming up with new names for those really, really long numbers.For those math geeks, it's 10 to the 21st power of 21.That's a Zetabyte.

(01:17):
It's insane. So with that, I guess the challenges that we see with all that data out there is you're juggling multiple systems as an RIA. You're entering data manually. You have multiple different applications, usually that data is sitting in some kind of silo, right? And trying to get it all together is really difficult. To make matters even more fun, the flows between those systems are very insufficient. So the information that flows from system to system is very limited. There's still a need for doing something manually or looking at five different screens as you're trying to talk to a client, connecting data from multiple systems. And if you think about what this conference is about, which is AI, the challenge becomes: how do you get all this data from all these multiple places and make sure that it's actually reliable? So something like an AI engine can basically determine and provide a consistent result and a consistent answer with the correct data every single time.

(02:22):
Because we're a regulated industry, it can be pretty challenging if the answer is different. "Hey, what do I have in my portfolio?" It better be the same answer every time. And most importantly, as you think about data and the future, data is kind of the new gold. You want to be able to organize it, clean it, and maintain it, and have a history of it so that AI can learn from it and give you insights to help you drive and run your business smoother. Also, very important is ownership of data. If your data is spread out across multiple different applications, how do you know that you actually own that data?

(03:08):
Because if you're trying to create something unique and harvest insights from that data, all that needs to be in your control. So whether you're an RIA, broker-dealer, if you're working with insurance or even real estate—we cover all these different verticals and we primarily help folks figure this part out. The challenge becomes: where do you start? How do you create this foundation, this environment where you can house all your data? And that's what Invent does. We built a data lakehouse that is specific for wealth management, which is essentially a big database with all the connectors that you need from all the applications that you use within your practice. And essentially, it's available in the cloud, it's available in a private cloud, and you have 100% data ownership with all the data that now you control. Data may be coming from multiple different systems, but now you have a historical record of every single piece of data from every single system recorded in your own data lake.

(04:20):
And on top of that, you start imagining what the experience would be for your home office, for your advisors, and for your clients. Very importantly, you start differentiating with the ability to build what we call a "super app," which is a combination of multiple different modules or micro-apps from different applications that you can bring together using the Invent Experience Builder to create your unique and personalized all-in-one that is specific for just your firm. Essentially, what this is called is an integrated data ecosystem. This is based on the Gartner study that was released in the beginning of the year where they talk about the future of technology architecture and where everything is going. I'm sure a few of us remember back in the day where we ran stuff on servers in the room next door, and then everything went to the SaaS, to the cloud.

(05:23):
The SaaS application, three-tier applications, all these things are kind of going away now. The future is essentially creating this data foundation, this data fabric, and then building apps and AI agents that leverage that same data storage. If all the data is in one place, you're not moving data from place to place. You're using it in one holistic environment. So your applications become, in a sense, agents that also do workflows or leverage that data to provide certain capabilities, analytics, onboarding, or whatever it is that you guys are working on throughout the day with that data. One other fun fact: Invent was named recently, actually this month, as the data lakehouse recommended by Gartner for the financial services industry. That is because of our focus specifically on wealth management and the ability to overlay all the entitlement structures and rep codes—all the fun stuff that you otherwise have to build from scratch to be able to have it out of the box.

(06:33):
If we look at the typical RIA setup, this probably looks familiar. I'm sure your individual map is going to be slightly different, but the reality is you've got certain integrations with certain tools. You've got your accounting probably sitting on the side here because your financial data from your business is not really connected to anything you're doing with clients, so you have to go to multiple systems. There's some integration somewhere, but it's not really organized. What we do with Invent is change this picture. Now you have centralized storage where all the applications and everything sits in one place, connected through a common data layer. It's an open architecture, so anybody could build to it. We don't charge anything for any of the vendors here in the room to build on Invent, adding to this data layer and providing custom applications to own your IP.

(07:31):
It's a completely open platform. What we see advisors do is innovate for themselves and figure out what things their firm can provide that are unique and different to differentiate from others. The way it works is really four steps. We start with essentially looking at the data, and that's what I recommend everybody do, whether you work with us or do it yourself. A very important thing is you really need to get to the point where you have clean, reliable data. The reality is, anyone we talk to says, "Yeah, our data's fine." And then we come in and see these rep codes on the custodian that are 000. I'm not sure what those are or who's looking after those accounts, but "I think Joe knows how to deal with this," right? That's what we're hearing.

(08:29):
And then, "Oh, the CRM—no, no, no. I don't use households from the CRM." For these types of accounts, they use Orion, and for this, they use Salesforce. It's really spread out. I'm telling you right now, there's probably somebody or a few people in your firm that really know where all the data skeletons are buried, but essentially that knowledge is not scalable for your business. Once you set that foundation and you organize your data, then you start adding some capabilities in our evolution step and bringing those applications into your experience for advisors, clients, and the back office. The last piece is "invention." If you have a great idea, now more than ever, it's really easy to build by leveraging AI technologies. Ten years ago, if you tried to build an application, you had to get servers and hire a bunch of developers. Now things are a lot more accessible using AI agents, and the programming process has increased tenfold in how fast you can build things.

(09:28):
That's another thing we provide on our platform that you guys could leverage. So, I wanted to talk a little bit about our panelists here today. On Invent, we have a hundred-plus different applications. Today we have Daniel from Finmate, and Finmate is an AI assistant for financial advisors. Again, think about deploying these applications directly on top of your data. We've been working with Finmate for just a few months to get their apps on Invent. Now it's basically a drag-and-drop, and you're able to use a note-taker application directly on top of your data with no integration projects needed on your terms.

(10:39):
Daniel, I just want to ask: with all this great stuff happening in our industry—if at the last conference we were here a year ago, I think it was like six AI firms, now it's like 60. That means every Tuesday there's a new AI company being born. With all this great innovation happening, and now people talking about segmentation of AI, how do you see this going into the future?

Daniel Yoo (11:22):
Yeah, thank you. Before we go into the future, quick background on us as a company. We launched in May of '23. We got on the Kitces Tech Map back in August of '23 as the first AI-focused company there. Since then, as Oleg mentioned, there's been probably more than 50—I think it's a few hundred at this point with all the VC money floating around. Where I view the industry going is platforms and SaaS are going to be quickly commoditized. There's no technological moat anymore. Because the cost of development is coming down so dramatically, I think the direction is a "done-for-you" customized solution that's specific for your individual firm. Embracing the commoditization—and I'm sure you've seen tons of other note-takers out there—I think there's about 30 now. It's kind of ridiculous.

(12:14):
We decided, hey look, it's a commodity, it does what it does. We are dropping the prices down to $39 a month, and we are expanding our solutions to custom agentic AI development, where we take you from consultation all the way to deployment on Invent as well.

Oleg Tishkevich (12:30):
Awesome. Great. Thank you so much. And I want to turn it over to Gary. Gary is our case study. In case you noticed, he's got a really amazing innovative RIA firm. I see a trend in the last couple of years where firms are focusing not just on assets under management, but expanding their value proposition beyond just traditional financial planning and investment management. You see larger firms like Hightower doing that where they have hundreds of billions of dollars under management and a lot of staff, so they start adding additional services. But what's interesting with Gary's firm, Helium, is they're able to do it on a smaller scale with a sub-billion dollar AUM. I would love to let Gary talk a little bit about your structure and the different business lines you guys were able to create.

Gary Russell (13:33):
Yeah. We actually started out running a broker-dealer RIA and tax service. Any CPAs in the room before I offend anyone? Okay, sorry. We started that and we kind of backed our way into running these other businesses that were integrated. We realized when it came to data, it was just a nightmare. If you have 3,000 tax and accounting clients as a small firm, and then you have commercial clients—we actually broke into what we're doing now because we had commercial insurance folks that were losing out on hundreds of thousands of dollars of commission based on like a $20,000 increase in premium. We'd have these multimillion-dollar relationships and people were losing those over a small increase. So we started bundling all these other services together to mitigate cost and keep those relationships in-house, which brought us to the data aspect of this.

(14:35):
In the process of doing that, once we broke off and started our own firm in Helium, our progression was developing our systems and processes to understand how to solve problems in our different lines of business. To the CPAs who raised hands back there: the first thing we started doing is going into firms and asking for the worst problems their clients had. We'd bring elegant solutions of how to solve them. Obviously, scaling that is really difficult if you don't have the right data set. Between us, when we were looking at creating our own or going to Snowflake—well, despite my partner having a tech background, he's not much of a programmer. As a small firm, you're like, "Where do you start? Where do you even begin?" Luckily we've had a relationship with Oleg for a long period of time.

(15:24):
For us, it was a really easy thing to plug and play. I think the really cool part of it is that with all of the applications there—if anybody has founded and started their own firm from nothing—when you're out there trying to build other advisors and teams, you're looking for the right people and the right fit. Unfortunately, what we found ourselves doing was looking at the right technology fit. Okay, what custodian are you with? What platform and CRM do you have? How does that match up, instead of looking at the people and making that match first? Now that we have a platform like Invent, it really democratizes all the other technology there.

(16:11):
What's exciting for us as a small firm that can at least innovate is having that data lakehouse there and being able to find opportunities inside these different lines of business as a really small firm. It's super exciting for us.

Oleg Tishkevich (16:25):
Very cool. Speaking of that openness, I wanted to bring it back to Daniel. If you can talk a little bit about commoditization and what you guys have done and how things are changing. I thought this was an interesting point about how we have 30 different note-taker companies now. We had one last year in that category. Being a veteran in the AI space—what, your company is 24 months old?

Daniel Yoo (17:02):
Yeah, we're the old ones, yeah.

Oleg Tishkevich (17:04):
Actually, believe it or not, that's like a grandpa in AI days. Can you share your thinking about commoditization? You guys dropped prices significantly recently. Just share what you see this market going to.

Daniel Yoo (17:22):
Yeah, absolutely. When we first came out as pioneers in this space and building out this little niche, I didn't realize how much venture capital money was waiting on the side to just flood in here. I thought this would be a nice little small tool, but at the time it was a premium product. We charged a premium $150 per month, but quickly, like Oleg said, the cost of development just cratered and, frankly, there were a lot of layoffs in the tech industry. The cost of developing new tools has gone down dramatically. This is why I think we're seeing 30 note-takers, but it's not just the note-taking space. I'm sure if you go out into the convention hall, there's a ton of these AI tools coming out month after month.

(18:07):
With such an embarrassment of riches, prices will need to come down and every one of these point solutions will become commoditized. Where the industry is going, I think, is with a data layer/data lake that all of these other tools are feeding into. Not only that, these platform-first solutions require the end user, the advisor, to adapt themselves to the tool, as opposed to the tools adapting themselves to the advisor. The big question on all the technologists' minds is: how do we get adoption in a very tech-reluctant industry like ours? Trust me, I was at TD Ameritrade for a long time. I was faxing forms back to our back office for Roth conversions only a couple of years before I started Finmate.

(18:58):
The direction we need to go into is building the tools for individual firms and how they do their operations, as opposed to trying to force the advisory firm to conform to how the technologists work.

Oleg Tishkevich (19:12):
Very cool. Speaking of that, let me bring up this other slide to talk about all these different things you have within your practice. Gary, if you think about your workflows, onboarding, and M&A strategy, how does your firm view those particular challenges and how do you guys approach it?

Gary Russell (19:36):
Yeah. Now that we have a solution for it, it becomes less about the structure and time it takes to onboard someone. I don't care what CRM they have or what note-taking tool they have, no offense.

Daniel Yoo (19:54):
It's a commodity.

Gary Russell (19:55):
But even when it comes down to the tax and accounting side—anybody that's ever been in QuickBooks or any of that software—it's like dinosaurs that are waiting to invent. If you can grab that data, it allows you to invent inside of it. From a standpoint of our firm, it's giving us the opportunity to find the right people and do it at a much quicker pace and lower cost. We can onboard those individuals who are the right fit as people, instead of focusing on what their CRM is or who their custodian is. It just democratizes that. It doesn't really matter anymore.

Oleg Tishkevich (20:36):
Gotcha. So streamlining those workflows, putting guardrails around processes, and then being able to adapt to whatever systems firms are using.

Gary Russell (20:49):
And to be able to do that in the same data lakehouse and ensure that our own systems are communicating to one another. We've had great conversations already about being able to query that data. To have all those systems as a small company and be able to literally by voice eventually query data and understand what's happening within the book of business—and then also build your own tools. There's some things we've done where we identify opportunities to help clients in QuickBooks and tax software, and for us, it's been incredibly manual up until this point. So we're excited about building some of those tools as well as some other fun stuff we're nerding out on the AI side, understanding how to extract that information and make it actionable.

Oleg Tishkevich (21:46):
Very cool. So you guys have an RIA, an accounting business, a—

Gary Russell (21:50):
We launched a private equity fund, and it's going really well. We are also partners on a group of guys who did an extraordinary amount of tax credits in a year due to technology. Even there, our CRM in the advisory practice has been Redtail for a long time, and we used HubSpot on the other group. There's all this data sitting there waiting to integrate, and the intersection of those data sets is what I'm really excited about.

Oleg Tishkevich (22:25):
Very cool. Gary and the team came up with a really cool AI idea that they're augmenting their management team with, which is pretty amazing. Stay tuned. Probably right now, what does it take to build new agents—four hours?

Gary Russell (22:45):
Yeah.

Oleg Tishkevich (22:46):
Yeah. So maybe by the end of this conference, we'll release it. There's some really cool innovation they've done from a management perspective, applying AI to various processes within their firm's management. What I liked about it is not just the operational use of AI, but even guidance and management of the firm. As long as you provide the right information, feed it the right data, and train it with your specific preferences, you can really automate a lot of tasks beyond just operational.

Gary Russell (23:34):
And I think the interesting aspect we were talking about—you mentioned they did a study on a judge.

Daniel Yoo (23:41):
Yeah, they did a study on a judge and the human element. They were tracking sentencing rates. What they found was if you met the judge right before they ate lunch, you were much more likely to get sentenced than if you met the judge right after he had lunch, because he was feeling good. Part of the conversation was removing that human element out of decision-making and getting it into a more objective frame of mind.

Gary Russell (24:07):
Yeah. Imagine being a small group and trying to make decisions, especially on a portfolio in private equity. You're making emotional decisions because you spent a lot of time on that. Mitigating that through the utilization of AI has been really eye-opening for us.

Oleg Tishkevich (24:35):
Yeah. So guys, next time you go fight that parking ticket, just make sure you schedule the appointment for after lunch.

Daniel Yoo (24:44):
Buy them lunch.

Oleg Tishkevich (24:46):
One other thing I wanted to bring up is that as we work with firms looking at the future of their practices, I think the common theme is: "let's automate this, let's put everything in line." There's a lot of anxiety about what's happening with AI. Are we going to be replaced? We all probably heard that a few years back with RoboAdvisors. I don't think human advice is going to be replaced, but if your firm is not leveraging this technology to streamline the boring stuff, you're really missing out. You're missing out on opportunities to service more people with quality financial advice.

(25:45)
Be able to lower the bar of the advice, be more efficient in starting to work with younger clients, or clients with less initial AUM. There are tools to automate estate planning or bringing in an expert to do that. Those human conversations are so important. People need to hear it from a human with experience and life spent in this industry. If you're able to update your tech stack and get it to the point where you can scale, you'll be able to service more clients. It's a win-win for everybody.

Daniel Yoo (27:13):
Absolutely.

Oleg Tishkevich (27:15):
Any questions from the audience?

Audience Member 1 (27:27):
I keep reading that the number one issue in the industry is integration. So how will this solve that?

Gary Russell (27:37):
I'll tell you how it's solved with us. To your point with Invent, when we come upon an advisor that had a different CRM, a different billing system, or a different custodian, that used to be a real problem. Now, for a small firm to find those folks, even though it's still a little bit of lifting, it's a quarter of the lifting that it was before. The timeline is way quicker. They can be on any of these other platforms and we can still bring them on board quickly. You can focus on the right people rather than the right technology.

(28:26)
And with the note-taking, I'm still a dinosaur—we're just adding the note-taking. I walked around with a Rocketbook forever. You write in it, it converts it to type, and you upload it to the CRM. This is a thousand times easier because it eliminates more of the human component, like when you get busy for two weeks and haven't entered any data. The human component of tasks waiting gets completely wiped out.

Daniel Yoo (29:08):
One quick note on the question of integrations: we've always been bootstrapped. Some of the valuations for these AI companies have been ludicrous. We realized quickly that to get access to these integrations, you have to be part of that network and "grease some palms." Working with someone like Invent has been really helpful because it centralizes the focus on innovation and development as opposed to political gaming.

Oleg Tishkevich (29:43):
Very cool. To add to that, right next door we have our Invent Village where we have some of our partners that are built on Invent. We're trying to promote this concept of integration and open architecture so you can adopt new technology without being stuck on something legacy. I would love for you guys to visit and talk to some of our partners. Thank you so much and we'll see you at the break for lunch.