AI, Risk Surface, and the Fragmented Advisory Firm — A WealthTech Reality Check with Adrian Johnstone

Most advisors are thinking about what a new AI tool can do for them. Fewer are thinking about what it does to them.

In this episode, Shelby sits down with Adrian Johnstone, co-founder and CEO of PractiFi, for a direct and practical conversation about technology risk in wealth management. Adrian has spent 15 years helping firms simplify their tech stacks. Now, as AI floods the market with point solutions and plug-and-play integrations, he's watching a new version of an old problem play out in real time — only faster, and with higher stakes.

He breaks down the concept of "risk surface," explains why the most dangerous integrations are often the ones advisors think are harmless, and shares what he believes is coming for firms that are moving fast without a real foundation.

But this episode isn't just about risk. It's about clarity. About knowing what questions to ask your vendors, how to think about your data before AI touches it, and what it actually means for a CRM to become a system of action rather than just a system of record.

If you're an advisor feeling overwhelmed by technology — or a firm leader wondering whether your stack is built on a solid foundation — this one is worth your full attention.

Topics covered:

• Risk surface: what it is and why every new vendor expands it
• Recent industry breaches — and what actually caused them
• DIY integrations, MCP connectors, and "vibe-coded" AI tools: the real risks
• What breakaway advisors get wrong about tech from day one
• The taker vs. giver theory — why advisors are rushing to AI right now
• WealthTech fragmentation: 100 solutions in 2018, 550+ in 2026
• CRMs evolving from record-keeping to agentic action
• Why "AI native" is mostly a marketing claim
• Lightning round: what to do this week, what to stop duct-taping, and where AI's biggest compliance opportunity lies

Hosted by Shelby Nicholl. Produced and edited by Aaron Sherman. Operations and Guest Coordination by Shelly Hadel.

Transcript

0:57

Welcome to the podcast, Adrienne. Appreciate you being here. Hey Shelby, I'm so excited to be here. It's going to be a great chat. I'm very much looking forward to it. Well, for listeners who may not know your full story, walk us through your path in WealthTech and how did you end up leading PractiFy? for sure. the conversation of does anyone ever actually have a plan or ~ a map for their journey? And ~ famously before becoming a founder or a practifier, I said the last thing I would ever want to do is tech sales. And so I definitely didn't follow that plan. But for a long time, I worked as a consultant to the industry. So I sat at the intersection of technology compliance and you know, the kind of evolving client experience ~ that people wanted. that was when software was moving to the cloud and, big scary things like that were happening. Nothing like what's happening today.~ And then coming into, to Practify as one of two founders, myself and Glenn, he really had the tech bent and I really knew the industry. And so it's been a ~ fun journey to to switch into the CEO seat over the last couple of years and really have accountability for both and to immerse myself even deeper into the tech. Was there a moment, you know, maybe that led you into this element of tech where you said, gosh, know, advisors don't need better technology. They just need better outcomes, better things that are happening with the technology. Kind of talk about your approach on that. Yeah, so it goes way back to when I was a consultant. And one of the most consistent messages was people like, I need another system to do this, and they need another system to do that. And ~ I used to use Excel as a simple example and say, use Excel to create a table of written text that you could do in Word. You're exercising 5 % of the functionality. And I would work with clients and review their existing technology. And the answer would be that they knew what their features were or what they do with it when they first implemented it five years earlier. But they hadn't kept pace with the technology. And to their mind, they thought the technology hadn't kept pace with them and their needs. And so it was often about helping them to use the tech they had properly and better and to exercise the full capability of those systems.~ That's where the value would be. You know, and it's an interesting time in the market right now in wealth, where we're seeing this kind of resurgence of hundreds of point solutions that do very small jobs. And I kind of sit on the sidelines a little bit now. And I think I worked with firms for 15 years to simplify their tech stacks. And now it's like a grab bag of candy where it's like, there's another system, another system. It's like, you can already solve this problem. You don't need these other systems. Yeah, I think it's so interesting. There's like this shiny new object syndrome. And then to your point, it is a big bag of candy. Hmm. And then I talk with advisors all the time. I had one tell me the other day, he's like, I know my tech stack is a little bloated. I know it is, but I really wanted to try these things. And so I find they're either deploying many, many, $20 here,$80 here, $1,000 here type of things, or they have an actual strategy. in that place. I talk about tech stacks with some of our build and RIA clients, and we talk about it as what's the leanest, narrowest, tiniest tech stack that we can have on day one? And then from there, you can go crazy, right? Like later on, if you want to bloat that thing out, you can. But maybe let's not do it on day one. What do you think about that approach? Look, I the day one approach is right. I'm not sure I want them to bloat it out because I think the piece that people Anytime. underestimate and don't think about at all is every new system that you add, whether it's AI based or not, it increases the risk surface. And they're looking at what they want from the system, but quite often they're not looking at the risk surface. How many places is my data now potentially exposed? How consistent is my data and what's the risk to the client experience that I might have? that's becoming so amplified in the AI era because of course nowadays I can build this myself. And it's not only the risk surface, but it's what am I building and do I understand? I wonder how many people who have built something in open AI or in... or whatever, from within the industry, the practitioner side in the industry, are tracking how many breaches of the code base there's been for those providers. Yeah, I think that's a really good point and one I really want to dive into it, but you used a phrase there, risk surface, that I'd love to hear you describe a little bit more on that and like, what does that exactly mean and how can an advisor firm think about their risk. surface? when you think about technology, you want to break it down into there's the system itself, the interface that you're using. There's the data structure and the data storage. So where is the data residing?~ And then there's how many vendors and what are their processes to control both of those things. sometimes we see advisors and if I use recent examples, it's Well, I want to grab this note taker tool and I want to connect it to my planning tool and my CRM and perhaps my portfolio tool.~ And what that does is it creates a big risk surface. So now you've got a record of the client, a partial record of the client forming in that note taking capability. You've got another partial record or a subset of the data being pushed to multiple other platforms through new connections. Each of those is data moving through a pipe and each of those pipes needs and management and control and all those sorts of things. so I think advisors or firms look at the connectivity only through the upside. So yes, integrated systems are great, Mmm. but the larger the number of vendors, the larger the type of integration, let alone the number of them. You know, the method of authentication, all of the the messy, scrappy detail that they don't want to know, the more of those there is, the bigger the surface area is. And they try and mitigate it a little bit with, at times, diligent surveys. So perhaps an annual survey, what do you use in this instance? What do you use in that instance? But it's often a bit of a box checking exercise. It's like, yes, I asked them this, they told me, but I don't actually understand it, and I've not delved into it at all. I'm trusting. Right. they, yeah, they don't understand it. They are trusting. How could, what should they be asking? They're doing a, they have to document that they've done some due diligence, but that due diligence process written is probably written by Claude too. How, what should they be asking? Yes. So I think that the first thing that they need to ask themselves rather than asking the vendor is, how can I achieve the commercial outcome I want to achieve? So the efficiency outcome I want, the user experience that I want for my team, and deliver the client experience that I want with the fewest providers possible. Now, in doing that, that's not to say everyone should use an all-in-one, because the trade-offs there are that you get one provider trying to be everything to everyone. so functionally the internal experiences is least good, but you want to try and say, if I've got 10 things I want to achieve, can I achieve those 10 things in 10 providers? Yes. Can I achieve those same 10 things in five providers? Probably. And so you're better in five providers. So that's the first question they need to ask themselves. And then With those providers that the first thing they want to do if there is a RIA is understand where the tech quality sits and they can do that in simple ways. They don't have to understand the technology. So is this something that someone built three months ago using Claude or is this something that uses the same technology that Merrill and Morgan Stanley and those guys use because They can leverage the fact that if those larger institutions have done the vetting in order to use the underlying tech that's there, that they can increase their trust envelope. The fact that they're ever gonna understand it as much. If they're using something that's newer or more emerging or less clear, then they really need to start asking questions about where is my data? How is it protected? And when they get an answer back, well, you know, it uses this type of security. They really should go to Claude and say, is this security robust? What are the risks for this type of data with the Like it's easy for them to learn a little bit.~ They shouldn't just accept, you know, Mm-hmm. an encryption answer without understanding what that answer means, just as they wouldn't accept an investment risk answer without understanding what the investment risk. I think that's really great guidance. And now that we do have things like Claude, we can go do a little bit of research and figure that out. To your point, what the big corporations are using is, it at least means it's gone through a rigorous testing program, and especially when it came to cyber. I can think about my, I spent 25 years in corporate America before launching Muriel Consulting and the background of vendor management and vendor evaluation in the large corporates is really deep, right? They have people that are just skilled in that aspect of the business. And so you can trust those providers a little bit more.~ I think the other thing that happens for many advisors or just even many entrepreneurs of any kind is almost an assumption that the people that you have built, that has built this new tool for you, that they have covered that. Like they know what they're doing, they've covered it. And so there's almost this like automatic trust in some of these providers, especially if you see great utility in the tool. They need to go deeper, right? So much and they need to understand what the provider is doing. for example, Anthropic dropped 10 new financial services connectors. And I immediately watched LinkedIn blow up with advisors going, I'm going to use this. I can replace this. I can do that. Now, Anthropic themselves say that these are not intended for end user use. These are for software companies to put tight rails around, to run rigorous testing around. they, are working hard to make them as good as they can, but they assume they're going into the hands of tech professionals, not someone who vibe coded a solution over the weekend that they're now unknowingly pumping all of their clients PII into a public model. And so, so for me, my god. I mean, I, I couldn't be more bullish on the potential of the technology, but I also feel pretty certain that we're headed for a fairly catastrophic breach somewhere. ~ Well, and we've seen some breaches already, right? I think it was, know, Mercer, EP Wealth, I think Anthropic had a breach, right? Like, we've seen breaches. We have, I think there's some things to unpack in those breaches. And again, this is where I think maybe there's an opportunity for the industry to understand more. So if you take those three that you talked about, so Anthropic had their code leaked. And that was the kind of big thing that everyone talked about. And that's a real problem, ~ of course.~ But they've also had malware that's popped up in there and all sorts of things that happened that don't make the mainstream headline because it's not that big scale. In the case of the more industry related ones,~ EP, Mercer, Pathstone, there's been a number of them in recent times, they weren't technology breaches. Each one of them was a successful phishing attack. So a user's credentials were breached, Mmm, okay.~ mostly. As far as I'm aware, and certainly I don't, I want to be clear to the listeners, I don't have the inside running on this. It's just, ~ it's being close to the industry. But as far as I'm aware, by a user giving over their credentials to someone they thought was a trusted party. So actually in each of those instances, Got it. Yeah. the technology worked just fine. The human failed. ~ And that's hard to protect from a technology perspective. Yeah. There are controls that you can apply that make that more difficult and all those sorts of things, just simple things like multi-factor authentication. But the reality is that the human remains the weak point in tech. It does. some of the phishing attacks, et cetera, are very, ~ they're very mature, right? And so we do see, even in our small business here at Muriel, we see people sending me contracts to sign and to log in to sign something. And I'm asking my my husband works with us in the business and he runs our tech. And I'm like, is this real? Is this not real? And he's like, no, that's not real. He looks at me like I'm crazy. But I'm like, well, it kind of looked real, you know, so I can appreciate how the human fails. totally. And this again is one of the things that I'm not sure that people think about. They get very excited by the potential of technology. And they forget sometimes that the bad guys have the same tech, but often more motivation to deploy it. Yes. And so you start to see in the industry, Yes, yes. for example, there's often been this view of, the way to test whether a client~ is genuinely asking for the thing that came in an email is ring them. Because you know what their voice sounds like. But voice sampling technology today makes it so easy for a cyber criminal to be mimicking the client's voice. And so you need that next layer. You need the keyword or the password or those pieces of information that are the next level of protection because the attack is getting the next level of sophisticated. And I think the point that I would make is,~ and, you know, Practified is built on Salesforce and there was a lot of~ probably six months ago about Salesforce having a breach. again, Salesforce wasn't breached at all. An integrated system was breached and that's that risk surface. So the, the challenge becomes that when you take four, you know, you take a five, ~ system tech stack and you bloated out to 10. You've not only doubled the number of vendors to manage and the fees you're and all those sorts of things, but you've probably more than doubled the risk surface. And that's something that I think is, people are a little asleep at the wheel. Yeah, you've said something in the past. You said something like DIY integrations are one of the biggest hidden risks. And I guess that's what you're getting at is it is increasing those risk service, that risk surface. And DIY integrations, that's things like, know, e-money to Zapier to Google Sheets. Is that what you're meaning by those DIY integrations? So it's a little bit that. It can be ~ even more aggressive these days because, of course, you've got ~ MCP servers popping up. You've got people using Claude to use their connectors. And so suddenly, you're starting to see people pushing and pulling data between~ systems. And as they move the data, they're thinking, well, it's not that many pieces of data, so it's OK. But they forget that in order to insert data into a system, you have to open a gap. And so you might only be inserting two pieces of data. Hmm. But if the system that it's coming from has been breached, those two pieces of data may carry with them a worm. And now you've got a worm in a system you're not looking at because you opened it, but you're worried about the source. You're not worried about the destination. And now you've put a pollutant into that system and now you've got a problem in a system that you'll detect. But again, for a lot of firms, particularly smaller firms or teams that are breaking away where this has all been centrally managed until the day that they form their firm, suddenly they're in an environment where they're responsible for those things and that gets pretty scary. Yeah, it is. And how do we bring down the temperature of the scare through maybe a process or at least a diligent conversation? What would you recommend to those breakaways? So I think if you're breaking away, one of the first things that you want to do, there's a lot of things to do. And so again, to those listening who are breaking away or have recently done that or considering it in the future, I don't mean that you have to do this exactly at the moment, but you should probably do it beforehand. Actually, just there's great online courses. Again, use. use some generative AI tools and try and inform yourself on where the risk lies. ~ Or if you don't feel that you have time for that, go to Claude and say, we're a team of six people managing 600 million going out on our own. What's the most mature risk aware tech stack? that we can put together for this RAI and then go through and read and understand and inform yourself on the content and then really drill those vendors and be transparent. It surprises me how few firms, because we deal with breakaway firms fairly often, it surprises me how few are prepared to come and say, look, we don't really know what we're doing when it comes to technology. Can you help us? Because of course we would help. Instead, they're like, yep, buy this, implement that, that's what I need. And it's like, you're already hard to get on the phone because you're trying to calm clients and move them with you and do those sorts of things. And now you just want this to tick away in the background. And because that's what they're used to. So I would say, inform yourself, make, understand that particularly at the point you break away, this is a full-time job for someone for a period. Yeah, yeah, definitely the setup is for sure and having, you know, good vendors, good partners all the way through and then using them, right? So, you know, when we are working on a breakaway into a build of an RIA, you know, use us, use the Schwab people, the Pershing people, right? The Fidelity, Totally. Yeah. Goldman's, you know, any of the custodians will have people who are experts also in the tech and how it integrates with their tech. And they can tell you also about the types of firms who are already deploying the technology that you're considering. So ~ that will help also kind of go back to that earlier part of our conversation about who's using what and what was their process for choosing it. know, one of the things I hear a lot about right now is just in general, we've kind of circled around this idea a lot already. but it is this idea of AI overwhelm, of technology overwhelm. Why are advisors, it does feel like that shiny object syndrome is coming up. Why do they want to adopt tools before maybe defining a problem to solve? Or how would you direct them when they're thinking about adding technology to their stack? Yeah, for sure. And I have a theory that answers, I think, the why. so because it's often advisors at the moment. And so over the last 10 to 15 years, there's been these trends that have moved through the industry. It's been very efficiency oriented. We've had a generation of advisors moving later in their career where adopting new technology, targeting the advisors has been lower. And that's moving into the next gen advisors who want new advisor oriented technology. But fundamentally for advisors, their world when they look at technology is one where the tech is always taking something from them. I want you to enter a note, I want you to type in a record, I want you to start a process, I want you to do something and in return you'll be able to access some information that's of moderate use to you but high use to operations, compliance, client service. management for reporting all those sorts of things. So advisors have really viewed a lot of the current technology, including CRM from our perspective, as what I classify as a taker. It takes something from them, but it doesn't give them much. And now what they're seeing is this big push towards AI technology, and it feels like a giver. Instead of, want you to put notes in after the meeting, it's like, hey, invite me to the meeting, I'll do that for you. Instead of, you know, come here and read information to prepare for a meeting, hey, hope you're having a great morning, here's your meetings today all pre-prepared. And so it feels like a giver. And so they're like, well, I finally, I want something that is giving back to me. But they're thinking of it, this is probably a massive generalization, but thinking of it selfishly. Because firstly, as we talked about before, that creates a risk surface, but it also creates silos because now they have a subset of data. that the rest of their team can't see. And so the client experience is now being fed out of multiple pieces of technology and bits of data moved between them, but not the whole set. And that happens because the technology vendors, the new giver technology vendors, they need to track the advisor's eyes because that's the retention strategy for value in their technology. But the rest of the business is kind of secondary. And you even see that in pricing. An advisor is $150 and their service person is 25. That's relative to the value they think that Mm-hmm. going to get. And so when you look at that model, you see the breaking apart, the refragmentation. Now what happens is the wealth industry is very complex. You look at the Kitsis wealth map and the FinTech map and you go back to 2018 and I think there was like a hundred and something pieces of technology on there. 2026, it's like 550 something. Yes. And that fragmentation is because tech vendors are saying, well, I know there's this pain point for this user and this type of phone. And so I can exploit that and build my niche. And then of course they get some traction. And so all of a sudden you see another vendor and another vendor and another And it's easy to pick on note takers, but they went from none. targeting the industry to about 30 in eight months. And of course they all fragmented in slightly different ways and they all do, you know, we do call coaching, we do better meeting prep, we do, and it's all these little add-on features that are all about giving back to the advisor, but they blind them to the fact that that shiny thing is creating this underlying problem in the business. And there's already now this movement that's starting. or in conversation that's starting where people are like, well, hang on, but now I've got an AI in seven different systems, but they don't all see the same data. Yes. So they're all giving me different answers. So now I've got this overwhelmed problem where, don't, do I ask my tax system this, or do I ask my note taking system that's morphed into something else this, or do I ask my planning tool this? And they're just getting the fragments. And so what we're seeing is this move back to kind of where people were 15 years ago, where they're like, well, now my data is all fragmented and it's siloed. we know that as soon as you have a data problem, AI runs on data. It runs on rules and it runs on data. The rules need to be structured for your firm and they need to live in the system that the AI is feeding on. And so does the data. And so the rules are now fragmented. the boundaries are broken and the data is fragmented. And so the problem is amplified by AI and not solved. It's really interesting to think about it that way, is that the problem becomes amplified. And it also is interesting to think about, like, we're going to see maybe a surge back from point solutions, back to more of an integrated approach. And certainly, like, CRMs are the heartbeat. I've said for a long time, they're the heartbeat of the practice. What do you see happening in the CRM landscape specifically, since that's your core business? for sure. And I think not enough is probably the top answer. And I think every advisor listening to it be like, yeah, what he said, not enough. But the reality is that systems of record, CRMs and portfolio management systems, those things, they've existed for a long time with a compliance focus. How do we store and protect? And they need to, How do we document that we had the conversation we were supposed yeah. to? That was the original piece of CRM, was documenting. ~ That's right. That's right. And if the SEC comes knocking, how do we evidence that it occurred? Well, we do that because we have a lot of structure and have a lot of rules and a lot of controls. Those things aren't less important. They're more important because the SEC and FINRA are very clear that if you are using an AI agent in any way in your business, the human level obligations attach. So you must be able to audit, must be able to cite and determine the logic. you've got to be able to explain it. can't be a black box. And so I think there's a risk coming there. But as CRM providers, as core system providers, our job is to move from being systems of record to systems of action so that the advisor can begin to feel that the system is giving them something as much as it is ~ managing the needs of the other users in the business overall. And so in PractiFY, for example, You know, we launched our PractiFi intelligence ~ capabilities in late last We're adding to those. We have some amazing stuff coming up ~ in a couple of But what we did was we said, well, yes, there's value in meeting preparation, but it needs to call on the entire corporate knowledge of that household, not just the last meeting. So let's prepare the advisor inside the CRM. Yes, transcription and summarization is valuable. So let's do that within the site with the inside the CRM. And then from that, let's take what happened in the meeting and use the DNA of the firm that lives in the CRM, the servicing model, the workflows that flow just as they want them to for the client experience that they want to deliver. All of that's already codified in the CRM. So let's build automation that pulls out the insights that require action from that transcription and automate the flow of it using~ tools within the platform. And so we delivered those pieces. We also said that evolving the system should be better. So it should be a more natural language experience. But that gets us to today. we, like every technology provider, we're worried about tomorrow and the day after and years from now. And so we'll have a robust, agentic solution available.~ for clients in the coming months. And that is where it starts to amplify the benefit. Where you start to get these autonomous voices that are analyzing and initiating and thinking and delivering insight to the advisor, to the client service person, to the operations team, to the compliance officer. ~ Where all of the perspectives are able to be catered to because it's in that system of record. It's not often just a point solution. advisor. It's really interesting, I think, this move from just note taking, right? That's been around now for a couple of years and now we're hearing so much more about agentic. And I was asked the other day by an advisor, is this AI thing really that big of a deal? I've got the note taker, isn't that all it's going to be able to do? And I was like, well, where I see it going and I see the RIAs moving here faster is in the agentic space. And also the... the larger ones are creating this kind of common data lake, data warehouse, know, database that you can basically pull all of your data from, this sort of data layer out there. When you think about, you know, the run-of-the-mill advisory firm, do you think that they will have the capabilities to create their own data lake and do their own agentic, or will they be better off kind of going through, you know, PractiFy and these other tools? Like, what's... What's realistic for them? So I think for a lot of RIAs, the question comes back to risk. do they have the capabilities? It's tough. You can buy those capabilities through consultants. But you're adding another, you're adding to that risk surface where you may not need to at their scale. But instead, what you need is tightly integrated core systems. So think about the fact that within an advisory firm, there's kind of two core There's the relationship data and the financial data, the asset data. And so you need a custodian who's the ultimate arbiter of the asset data. You're probably multi custody if you're an RIA, so you need an aggregation tool there and that's your out of power or your ride or whatever. And you need a CRM because that's the custodian of the client record. And so you've got those. If those systems are well integrated, then you can look to those providers for the agentic execution. they will, agents, I think people don't quite understand what they are. They're not magic and they are computer capabilities that are designed well and execute well. And they want good cross system capability. And this is part of the vision that they have where they have a data lake and then they build it on top and they say, it's going to go to all of the systems. But in reality, because there's so many checks and controls, is this the right person? Should this actually be allowed to happen inside of a set of rules? What it really is, is one agent talking to the next agent. And that comes down to the vendors working well together. And that continues to be probably the source of friction and frustration from an industry, is how do you get vendors that work well together? And I think for the RIAs there, the way to... to move in a stronger direction is to work with vendors who have those long established partnerships ~ and to understand that right now while the vendors are running hard, those who work together and have for a long time already have that connectivity. So connecting the next capability is a smaller lift. You've also said in the past that really technology should be invisible to the client or to the advisor in some way, right? Like it's really about the business and how do they want to do, how do they want to run their practice and deliver to the client and tech should be invisible. What is, tell us a little bit more about what that means. when you think about an end investors experience that they have, tech isn't invisible if they have three portals to log into and they can never remember the password and they have to ask every time. And I think for a long time, there was this big push is like, well, we want to make the advice firm ~ more efficient. So we're going to push. Form-fill technology to the investor and make them fill it in so text not invisible now because they're filling it in You've just given them more to do you just change who does the task? Whereas I think and this is where AI can help a lot is if a client is being on boarded at a firm and The advisor is hey, you know understand from our conversations you had a financial plan With your previous advisor. Why don't you just send that over to me? and your last tax returns and any other financial information that you've got. And so to the client, they're just sharing that information. They're either doing it in a portal, hopefully, or in a worst case, a secure email. ~ But when that information comes in, you then want it to be invisible on the advisory firm. And today it's not. So it's a high friction process. They get all those documents. Someone has to go through them, type in manually all the bits of information and try and piece it together. And so an invisible tech solution is going to read those documents when they land in SharePoint and say, hey, here's all the information you need to know about Shelby. Would you like me to update the CRM and push it to the planning software? That's where the tech starts to work for them. That's where it starts to slide more into the background. And I think for advisors, that feels like that acceleration got there very quickly. when you look at the conversational interface, the chat interface, where the question is, who crossed my clients haven't I spoken to for a while? When that isn't, will I need to build a report in a report builder that filters on my clients and asks for them by location, by asset type, by last contacted date, or was it an email, was it? You know, there's this whole thing that happens and it takes four days because the advisor doesn't know how to build the report. So it gets queued up for someone in operations versus them being able to go to their phone and either by voice or typing into a conversational interface and say, give me a list of the clients I haven't spoken to or interacted with in the last 30 days, 90 days, whatever. And they're instantly returned a list. So in that instance, the tech has become pretty invisible. The report building has become invisible. It's still there. The technology is still accessing it. But we're allowing the human to interact in a human way with technology instead of making the human have to interact in a technology way with technology. Yeah, it's really powerful and it is, it does feel like it's a sudden and all at once sort of change. And I think any of us that are, you know, thinking about tech and looking at tech where, at least for me, I should just speak for myself. I feel constantly behind, even though I'm doing things with tech that I know are reasonably advanced, I still am feeling like, my gosh, there's so much more for me to learn here. I need to set aside all these days of time in order to learn it. I like your idea of instead, like really picking and relying on your vendors to bring that stuff to you. It's going to be well developed. They're going to have done that work for you. That's right. And look, you're probably right to feel behind, but you shouldn't feel alone in that because I run a tech company and I feel You wake up every day to find out what's new. And ~ we invest whole humans into keeping on top of it. ~ And the reality is that it's a torrent. And making good decisions in that environment is hard. And this is where I think there's a concern and the risk conversation comes with, well, yeah, I like what they said. So I've, I've signed up because it's not expensive. So I've signed up for this $40 piece of technology and it says it needs to connect to my CRM and they know how to do that. So I've said, yes. And all of a sudden you've now exposed the largest source of client data that you have to a technology you maybe don't really and you haven't vetted well and you. don't know what they're doing to look after your data. Yeah, I think that's the biggest piece there is that you haven't really vetted it well, right? You sort of are trusting that it's,~ that it has done the sufficient work behind the scenes, but you haven't checked it. It is interesting, I think, with RegSP and a few other things here in the US, we're going to see more and more conversations around cyber risk coming up. And so it is something that we're talking already with ~ our advisors and certainly the compliance firms are talking about it and whether or not you're in compliance, having a good vendor due diligence process, et cetera, is something that every RIA should be looking at. One last major question I want to ask you is, is there's a lot of talk in the industry, in the tech industry, about AI native and that AI native companies Mm-hmm. are somehow suddenly so much better than companies that have been out there What's your take on AI native? Is it important or not? So no, think is the very short answer to that. But I smile because if you go to the sort of people making those assertions, getting a common definition of what that term means is almost impossible. But I think the thing that should matter, and what's being intimated in there is that the experience, the functionality within the system, is delivering an AI-centric user experience. So agents are in the system doing work,~ that a conversational interface is a part of it, not bolted onto it. ~ And certainly if you're a provider like PractiFy, we predate AI. So ~ we would be in the bucket of, you can't be AI native. But of course, we're going through an exercise of rebuilding the entire product to be~ what, what I think people are confusing there is if the system is new, therefore it's AI native, you can still build a system today that's not,~ and you can rebuild the relevant components to a system that pre existed So it's not a non lineage problem.~ I think it's more that they want an AI native experience. They want to interact with technology and they want agents that are doing things for them. But as we know and as we see, the AI needs guardrails, it needs structure, it needs rules, it needs the compliance framework, the client experience that a firm wants to deliver, and it needs structured data. So I would say to firms that instead of worrying about if something is AI native or genuinely can make that claim, what they need to do is think, how is my data organized today? How will it be consumed by this piece of technology and what will it do with my data? Because actually AI is just going to amplify the gaps and lack of control in data. that's great. Really appreciate this guidance today. I feel like it was such a needed conversation to kind of remind people of the risks that are out there, to remind people to do solid due diligence on any of their tool set that they're deploying. One of the things I love to do at the very end here is just have a few kind of really quick hit kind of lightning conversations, right? So just a couple of these. First one is... What is one thing that advisors could do this week to reduce their own tech overwhelm?~ They could not purchase a new system That's great. Don't add to your tech stack this week. Don't buy the shiny thing. What is one thing that you wish that advisory firms would stop duct taping together?~ Data. So I think ~ there is this tomorrow's problem issue with data. So we'll get to it tomorrow. Yeah. Stop doing that. Actually invest the effort today to understand the data you have and the holes that are in That's great. And what's one belief about AI that you think more advisors are going to come around to eventually? I think they're to come around to the fact that it's actually the biggest area of Neotone value is in compliance. Whoa, okay. Well, gosh, that's a whole other conversation, but really interesting. ~ And then We talk a lot about living an off-script life and what that means, a life that's full, a life that is entirely scripted by you and not just following the rules that sort of society lays out for us. What's a way, Adrienne, that you're living an off-script life today? I live between two continents that are far away. So think for most people, that's pretty off script.~ So yeah, Yeah, I think so. maybe I'll just leave it there. like, live between countries that are far away. Adrienne, thank you so much. This was exactly the conversation that we wanted to have today. And right now it's really about more, not more tools and more features, but really about clarity and remembering the risk, et cetera. So appreciate your reminder. To our listeners, if your tech stack is feeling chaotic, if you're overwhelmed by AI, if you're duct taping things together, just to get through the week. This is your sign to maybe rethink that foundation. Adrienne, thank you so much for being here. Thanks for having me. It was great fun. All right, to our listeners, we'll see you next time.