The AI Playbook for Canadian Wealth Firms: Bridge the AI Execution Gap or Get Left Behind

[00:00:01] David Kitai: Hello, and welcome to this WP webinar, the AI Playbook for Canadian Wealth Firms. Bridge the AI Execution Gap, or Get Left Behind. My name is David Kitai, I'm the Senior Editor at Wealth Professional. Just a quick note on housekeeping before we get stuck into this. There will be a Q&A session at the end of the webinar today, so I would encourage each of you to add your questions to the Q&A box that you can find at the bottom of your screens. I think today's webinar is particularly important because it deals with a subject that has defined market performance, business investment, and broader social discourse for years now. How are we going to use generative and agentic artificial intelligence? Today's webinar will tackle that question for financial advisors, outlining the playbook for effective AI implementation. To take us through this playbook, we are very lucky to be joined by John L. Connell, the founder and CEO of Focal AI. John, please, take it away.

[00:01:01] John L. Connell: Super. Thanks for having me, David, and appreciate everybody jumping in on today's webinar. As David mentioned, my name is John, I'm the CEO of a company called Focal AI. We are an AI meeting assistant for Canadian advisors, saving folks 10-plus hours a week through AI note-taking performance coaching, data entry automation, across advisor tools like your CRM, Conquest as well. We work today with firms and advisors that stand across Manulife, Financial Horizons, Hub, Investia, IA, Desjardins, IPC, IG Wealth Management, among many others. Additionally, we help organizations set up and deploy Agentic services. Using AI agents that self-manage workflows end-to-end with your data, with your tools as context. And, the last few months, we've been traveling, really, across Canada, everywhere from Ottawa to Calgary, Vancouver, Toronto, we have a home base there, Niagara Falls as well to talk about AI and best practices that we've seen in successful implementations, along with where the technology is at today, and what's coming next for the industry with Agentic Workflow Automation. So we'll be chatting on a few of these topics here in this conversation.

[00:02:21] John L. Connell: Now, as we kick off, we've got an initial poll question for everyone, just to get things started here. And if you will, please just respond to the poll. Very curious in terms of context setting. This will give us a sense for where different firms are at, and at what level it might be helpful to drill into some of the topics in today's conversation. And while you're looking to that poll question, I'll go ahead and jump forward as we…Get the responses pulled together here. So, diving in while the poll question responses are coming in, it probably comes as no surprise to folks in this room especially, that just about every Big Five Canadian bank has named AI as a top strategic priority. It is the fastest growing, fastest adopted technology of all time. And for context, it took the internet roughly 13 years to hit roughly 800 million users. And, ChatGPT landed that in just a mere 3 years. Now, yet, roughly 95% of organizations are not generating what they perceive as value from AI. And you probably saw the recent BCG report as well, stating that 60% of enterprise AI deployments actually fail in the real world. So when we ask ourselves, okay, what exactly is going on here, right?

[00:03:52] John L. Connell: It's effectively the implementation paradox. We're seeing that some firms are effectively deploying AI, they're seeing the ROI, but the reality is that most are failing at implementation, because they're misunderstanding how to scope or how to deploy the technology effectively. I saw some of the poll responses just now, and it looked like roughly a third of folks are trying to figure out what to do with AI. I saw another third are starting to deploy it within practice, and then the spread from there, with higher adoption internally already is being realized, which is great to see. Now, I'll be sharing here for those that are just trying to get started, initially, a few frameworks to navigate AI implementation, just so that your firm doesn't get caught in that skeptic's camp, right? Execution, ultimately, is the great separator between a successful AI implementation and one that just doesn't quite work out as well.

[00:04:48] John L. Connell: So, taking a quick minute to look at the macro themes, so we can understand the importance of bringing AI into your practice, why are we here, right? At a very high level, the $1 trillion great wealth transfer is happening, and so one of the questions that advisors are asking themselves is. How do you capture even a slice of that opportunity when half of all next-gen investors are planning to transfer their advisor? Right? Client expectations are on the rise, firms that don't innovate will be moving slower, serving fewer clients. One of the big upticks for AI is that increase in client capacity. And successfully adopting AI is going to be a catalyst moment between those winners and losers over the next decade. So if you're a firm looking at this opportunity, right, today. You probably realize that every advisor at every firm, unfortunately, has capacity limits. Advisors today are spending 20-plus hours a week on work before, during, after meetings that does not generate revenue.

[00:05:49] John L. Connell: So, if advisors are capped at spending just, you know, 17% of their time speaking with clients, and the great wealth transfer is going to be the opportunity of the decade, literally, if I'm in your shoes, I'm asking myself, how do I double or how do I triple the number of client conversations that my producers have? And the second level consideration is how do you improve performance across the board to up-level conversion of your prospects, the outcomes that you have with your advisors? Is it possible to operate like a top 10% producer in the industry? So…That's really where we come in, and we help organizations, both save a bunch of time, but also looking at problem statements around how it is that advisors are performing, and visibility into that performance within a given firm. So the name of the game here really is time to value when it comes to technology deployment. you, as a firm today, have existing tools, you have processes, and one of the questions you're likely asking yourself is how to seamlessly overlay AI to accelerate your practice, right? How do you set your firm up for success in the AI era? And so that's where we'll start the conversation today, with the AI playbook.

[00:07:11] John L. Connell: So this playbook really pulls from best practices, from new technology deployments, advocated by a number of consulting leaders across MBB, that's McKinsey, Bain, BCG, some of the leading consulting practices today. Your starting point for an internal initiative is really around that opportunity identification, right? Understanding what needs to be figured out around the problem statement that you're trying to solve for then identifying what work must be done to ensure that you can tackle the opportunity, and the scoping of work allows you to determine the work required, and then decide whether it's even worth prioritizing. The last step here, being pulling the trigger with a potential path forward.

[00:07:55] John L. Connell: And just for some personal context, when I was back at Microsoft, we used to use this process to acquire, to integrate new technologies, LinkedIn, Minecraft, a number of others. And it's particularly applicable whenever you're deploying technology, or trying to identify, you know, how do you think about an AI partner for your business? So, when evaluating the opportunity space, it's incredibly important to be asking the right questions as you navigate this process. And the right second-level questions around magnitude, specificity, constraints. If you take, in our industry, note-taking as a case in point, maybe your team identifies that it's a problem with solving, you must then identify how you're going to measure success. Right? Do you just need to hear good feedback from a few advisors in a pilot cohort? Does it need to integrate with your tools or understand advisor-client terminology? Should the tool also offer specific features? Maybe performance coaching? Maybe syncing data with your CRM or your planning tool?

[00:08:58] John L. Connell: Then figuring out what types of guardrails apply. Think time constraints, who should trial, how are you going to onboard and roll out a solution internally. Having an effective process and knowing your decision points is really a prerequisite for a successful AI implementation. So, if we take a quick look at some of the considerations here, right? On how do you make AI deployable inside of your firm today, the first piece is around data readiness. So, is your client data accessible? Is it connectable, right? Are there APIs? Do you have companies like Focal AI building AI overlays across different systems and tools for you? Do you have your workflow understood? What changes in an AI-native world? And what work gets done by humans relative to the system. Ideally, this is an empowerment tool that gives people time back and allows for them to spend more time with their clients. And then lastly, the decision governance, right? What is your, sort of, go-no-go state for trialing or for implementing AI within your practice today?

[00:10:05] John L. Connell: Now, this next slide, this is more in tune with some of the enterprise thinking. Here, but once you're structurally prepared for AI, one of the first questions that executives ask is whether they should build versus buy. This is oftentimes the first of a couple mistakes we'll speak through, that we're seeing within enterprise deployments today. Unless you are a tech firm and have a team of top-level AIML engineers that have a skill set to build towards Agentic workflow automation, we typically advocate for focusing on building deeper client relationships, growing your book of business, right? Empowering your firms. while leaving the intelligent robotic process automation to AI partners that have already built these kinds of systems or platforms, or work with partners that have platforms that you can build on top of as your AI layer.

[00:10:53] John L. Connell: The other piece that we're seeing, in a couple of scenarios play outis, the marketplace model versus a direct AI partnership model, within larger, larger firms. And now, the takeaway here is that if you're trying to experiment with AI one-off, or letting advisors pick what they want, your firm, as a larger, say, broker-dealer network, group of advisors. You're going to be dealing with multiple vendor issues across due diligence, integration requests, wasted time ramping up on training different pockets of users. You'll have internal conflict as well, about what to do next. And even worse, we've seen scenarios where advisors unlock, like, 10% of the value, because they're trying to self-navigate AI with something like an off-the-shelf, general-purpose tool. And then they determine that AI doesn't work. Because they're using bad prompts, or they're simply adopting a, you know, tool that is not necessarily purpose-built for their workflows.

[00:11:53] John L. Connell: So…organizations at scale will unfortunately fail to unlock the full value of AI if they don't control the deployments and onboarding top-down. One of the benefits of choosing a single partner approach is that you can then go deep on partnering for data context, which then allows for your firm to deploy agentic automations. That's where the real value of AI becomes uncapped, because it's basically Compliantly managing your workflows for you. So that, of course, brings up the question how you think about partnership, right, for AI vendors. Really, if you go down the list here, security and compliance, absolutely paramount. This is table stakes. The number of times that we've seen advisors adopt a tool that isn't compliant for financial services is pretty shocking. There's a lot of folks that are actually, right now, literally copying and pasting client information into ChatGPT, which is a big no-no on many, many levels, or using a tool that You know, is sort of a general note-taking solution that might not be fully compliant for the industry.

[00:13:00] John L. Connell: The second bucket to speak to here Is making sure that the engineering expertise that you're selecting for is deep. So in our case, our CTO had started his career at a broker-dealer that was bought by Goldman Sachs, and then has since built out systems for RBC, for GP Morgan, was an early employee at DocuSign, Chainalysis, so having that wherewithal around what the enterprise standard and the expectations are for how you build compliant infrastructure. Super important. The last piece here is around the Agentic roadmap. Really, it's more than just finding a point solution, especially if you're a larger entity, you need a partner to help not only unlock some of the access to data unification, but then be able to deploy data entry automation, offloading the clicks across, you know, multiple tools, and really tying together workflows end-to-end. That's where the real value, again, for Gentic AI is evolving.

[00:14:03] John L. Connell: The biggest risk here is really standing still, though, right? Not taking action while peers around the corner are really effectively deploying AI within their practice the compliance piece, this is super important. Making sure that you're working with a vendor that has a, you know, Canada-based data center. In our case, it's a Toronto-based Azure Microsoft data center. They need to be sure that they are keeping Canadian client data, you know, in Canada. We also really advocate for a few other pieces here as well. The storage of audio or video don't do it, those are artifacts. An AI partner should be looking at stateless AI architecture and ensuring that, PII, or personally identifiable information, your client data, isn't getting trained on or sent directly to third parties that might be training on that data. PIPEDA Aligned Consent is one that we've seen, even within our industry, a couple of organizations just miss the boat on. In our case, we use visible meeting bots to help your advisors stay onside of PIPEDA's implied consent requirements, so that mitigates the risk of some of those fines up to 100K CAD,

[00:15:19] John L. Connell: There are, for example, botless providers that can leave an organization exposed to some of the legal ramifications if they do not explicitly ask for consent. So we have a number of capabilities built into our system to help mitigate those risks. And of course, never storing audio, video, stateless models are a must. The last P-Series are SOC 2 Type 2, right? That's, at this point, pretty table stakes for enterprise software. So, poll number two! I'm quite curious to hear, what is the biggest blocker to AI adoption at your firm right now? Options include compliance, lack of executive sponsorship, fragmented vendor landscape. And just change management, advisor adoption concerns. We'll give folks a couple of seconds to… Fill that one out.

[00:16:21] John L. Connell: And we'll take a look at the responses shortly. As those poll questions get answered, we'll move on into the next slide here. So, AI is really designed to overlay across your systems and tools so that you can take action with the context from your data. If you are deploying it effectively, ideally it's not a rip and replace, right? This is purely additive, and a net new, really, for the modern advisor stack, especially when you're seeing peers, you know, save hours or elevating performance with AI in the mix. And pulling up the number from the poll here, very interesting to see roughly half of folks on the call are sharing that compliance is that big blocker, and absolutely ensuring that you have an AI provider that goes deep with your firm, your systems, can make sure that they are deploying, you know, efficiently and compliantly. That is a must.

[00:17:22] John L. Connell: Some of these other pieces here, it looks like fragmented vendor landscape came in at 22%, build versus buy confusion. Yep. Understanding, you know, what are the problems, the pain points that you're trying to solve. Start there, look at vendor selection, and then optimize for your guardrails, your requirements as a firm. I'm ensuring that those compliant stamps of approval have been granted. Executive sponsorship and advisor adoption, that 14-16% range. Very interesting to see. Appreciate everybody filling that one out. Looking at AI deployment one of the questions we oftentimes come across is whether generic AI, like a co-pilot, is good enough to do what they want. And the reality is these systems might be able to do bits and pieces of what a purpose-built solution could do, right? It might be able to take notes, it might be able to build out an agenda, that's like 5-10% of what something like a purpose-built platform would be able to do. The bigger challenges are that these general-purpose solutions miss the context of advisor-client relationships. They oftentimes miss industry terminology, and they certainly don't integrate with industry-specific tools if you're using a Maximizer, a Conquest, a EquiSoft, right?

[00:18:39] John L. Connell: It's oftentimes purely linear versus Agentic AI in terms of capabilities as well, so you can query it for a response, but it might not be able to offload a series of clicks across 4 separate tools, or automate your KYC or onboarding process for you. And it generally comes with a security gap. Different tools have different standards here, right? For the industry, but if you've got personally identifiable information exposed, if you're using public URLs or public API calls, there's a lot of failure modes that we've come across. And just to define for the folks in the room here, generative AI you could think about as just generating copy, so that's the agendas, that's the notes, that's the tasks, right? Agentic AI is where you get into automating the work. It is actionable. It is taking the clicks for you, it is, you know, inputting that data into different systems, it is self-navigating compliantly on guardrails with human-in-the-loop processes, how it is that your work is getting done.

[00:19:40] John L. Connell: So, if we ask ourselves how AI advisors, excuse me, how AI saves advisors time and elevates performance, I'll pull the curtains back a little bit on what it is that we're doing at Focal AI, just for some visibility for the crew that has joined us today, and then we'll see if we can walk through Super high-level demo before circling back to discuss what's coming next for AI in the industry. So, today, before every meeting, advisors have to flip through their old notes, their CRM, to determine what to talk about with a client. Now, imagine if you had a 360-degree overview of a client, right? An agenda, a pre-meeting email, a full view on the client that is in your tone, in your voice. And then Focal AI is able to generate this in seconds. Instead of meticulously taking notes, you know, we're looking at compliantly taking them for you, and then intelligently extracting those tasks. The summary is so that you can sync them directly with your CRM after client calls planning tools as well, filling out 400-plus fields directly into Conquest through our AI overlay. I'll show you shortly what that looks like.

[00:20:55] John L. Connell: The AI meeting assistant will also craft post-meeting emails, right? Allowing advisors to ask questions of their clients, having context of their CRM. The power of AI, gets unlocked when you get into the Agentic Services piece as well. Right? How do you then deploy AI agents to offload even more workflows? Because it has the context of the client-advisor relationship? So I'll see if we can do a quick walkthrough of Focal AI. I'll show you our Conquest overlay, our AI agent that sits on top of other tools, and then we'll come back to the presentation here. So, changing gears over to the high level, in our product today, once you've been able to integrate your CRM, your calendar, your video, your conference tools, right? You can think about Focal AI as offloading your entire client management process, end-to-end, before, during, after meetings with your clients.

[00:21:53] John L. Connell: So if you get into some of the pre-meeting workflows that we're able to self-navigate, for you, we're pulling context from past conversation history, think household summary, location, salary, their financial goals, what you discussed as action items in past conversations, both for the advisor as well as the client. The last couple of meetings, you know, what were those summaries? What were those takeaways? And then what are conversation starters that you can bring into play in that upcoming Back and forth with your client. Things like, what was their dog name, or what school did their kids go to? We also get into agenda generation, as well as client email generation that is fully customizable, personalizable, in your tone, in your voice, so that you have full control over how it is that you follow up with your clients and reach out to them before and after those client conversations.

[00:22:50] John L. Connell: So you can customize, you know, the different variables. For example, that get generated, the agenda items, or the, key financial terms. You can also ask anything of the agent that sits on top of Focal AI across any screen, to take action, to ask questions, if you wanted to ask what was…his salary, for example. Focal AI is basically pulling in context from the CRM, from client conversations, any other connected data resources as well that we might be partnering with, so that you effectively have at your fingertips your client data. You could think of this as, like, ChatGPT for client data across anywhere on the web. So if we get into the conversation history here…let's say we just had a client conversation with David. We're extracting the dates, the times, the numbers, of a given back and forth with the client, right? Extracting areas like life events, financial goals. If you discussed insurance or estate planning, all of this is being captured by the tool.

[00:23:55] John L. Connell: And it's all also fully customizable, so you could hypothetically come in, both for summary templates as well as email templates toggle on and off different types of output that you want to see the AI generate. You can also bring into play additional custom sections that you might want the AI to generate. Or you can build out entirely new summary prompts, based upon how you speak, what sections are important to you, what you want Focal AI to capture, and then what specifically do you not want it to capture. We'll also then extract the data points associated with the conversation here. You know, areas like retirement, or what age of retirement, or really anything that's associated with suitability, client profile, or financial planning at large.

[00:24:42] John L. Connell: Now, if you're an advisor, you've got all this great context, right? The meeting summary, the data table, now you're trying to figure out how do you update your different tools, that you use day in and day out today. Focal AI allows for advisors to push meeting summary directly to CRM, push their tasks that get intelligently extracted to the CRM, generate email follow-up, similar to what we showed earlier in the pre-meeting flows. And if we just get into that to show you a quick example, you can select a client. you can make sure that the meeting summary lines up, you push the CRM here at the bottom, and then that shows up in the object that you've already determined is necessary on the contact within your CRM of choice. So, in Salesforce, right, it just shows up in that notes and attachments section over here. The last couple of pieces here I'll speak to on the core product the coaching features. This is a scenario where we are working with the leaders in behavioral finance. We've got the science-backed approach with tens of thousands of assets that deliver on direct, personalized insights. What did you say? What did the client say in a conversation? And then, how is it that top 10% producers are engaging in that moment?

[00:26:00] John L. Connell: And very importantly, because we're using science-backed research here in partnership with PhDs over at Harvard and Stanford, we're able to directly align on how it is that you can really, truly elevate your performance and better prospect, better AUM conversion, we're seeing deeper client relationships as a result of the coaching feature here. 30-year vets telling us they didn't realize that they were, you know, missing out on how they could be going deeper around retirement, or missed opportunities that might have been pulling from past conversations. We also have the speaker analytics piece that allows for you to see the talk time, you know, what went well, what went poorly throughout the conversation, so you've got a direct retro. And then that Ask Focal Anything, ability to come in and directly either edit the meeting summary, you can ask, again, client questions here.And the beauty of this is that it really sits across your entire browser. So, if you're in email, for example, and you wanted to bring in the Focal AI agent, you could basically come in, you know, select an agent, ask what was his financial goals.

[00:27:13] John L. Connell: And Focal AI will be able to help assist, whether you are working inside of your CRM, working inside of Focal AI, or if you're just trying to craft an email on what his financial goals are, and bring that information into an email draft. Now one of the most powerful features with Focal is actually the AI overlay. So, it looks like I'm gonna have to quickly log in, to our Conquest instance here, but, Focal AI sits on top of other types of tools. And so, the AI overlay effectively automates the work to be done directly within the tools that you already use today. So we can come in, we can fill the form, Focal AI will identify all the fields that we can accelerate work on. As you can see here, as I'm talking, it's already filled in all these different, data object fields. You can also see what might have been missing. You can copy this for later, so you can keep tabs on it, right? And then, as you're navigating from one section to the next, because we're operating on guardrails, we know exactly what data needs to go where within the tool what work needs to be done across other advisor tools, and then we are offloading the data entry, we are offloading the clicks. This is where we're getting into intelligent robotic process automation. This is agentic work that we're building for your organization.

 

[00:28:38] John L. Connell: So if we move now back to the core, presentation, we'll see if we can…Talk a little bit around what is coming next for workflow automation. One second here, folks, I'm just gonna swap over to the other slide deck. Excellent. So… oh, I see a couple of questions here. Let me also respond to these. Somebody's asking, what is the strengths and weaknesses you see for Copilot, for financial advisors. When it comes to general-purpose AI, to the earlier slide, let me actually jump back to that other slide for you, Mark, so that you can take a quick peek at some of the bullets here. Generic AI will typically fail for a few reasons. One is that it might be missing the context of the advisor-the-client relationship.

[00:29:36] John L. Connell: Two is that it doesn't integrate with tools, with systems. Oftentimes there's additional work that needs to be done there. Some of the overlay capability that I just shared, like with the conquest planning, what we're doing is extracting more than 400 different data fields from client conversations, and then helping you auto-populate that across other systems and tools. Copilot wouldn't necessarily do that. The performance coaching piece, for example, right? That's not part of the Copilot platform. And it's really today more of a general-purpose solution set versus something that is vertical, purpose-built for financial advisors. And then the next question here, can it integrate with financial planning software like Conquest and Snap? Hopefully that Conquest overlay was helpful, just to give you a sense of what we do there. The next question, Dan, what CRM systems do we integrate with? There's a long tell. Salesforce. Equisoft, Maximizer, you… if you're using a Canadian CRM or a US-based CRM, we probably have an integration for you many different flavors of Salesforce as well. And, actually we can show you what that looks like, setting it up inside of Focal, after this conversation, if it's interesting. But it's very high level, a, overlay where you can basically map the objects, and just GUI interface, determine what it is that you want your data to be piping into what fields within your CRM.

[00:31:07] John L. Connell: Super. So, I will come back to the rest of these afterwards. Let me, actually fast-forward and just finish up the rest of the presentation here. A lot of great questions already in the, the Q&A, though. So the, really, next steps around where we're moving, right, as a business and then as an industry at large, is towards multi-step complex workflow automation, as well as building personalized recommendation engines. So you already saw how we're automating work inside of Conquest, for example, by automating that client fact-fill, right, household fact-fill. For more than 400 fields. We can do this across other tools, including bespoke systems for enterprise relationships. And this means stitching together different actions, including data entry, automation of clicks across advisor tools, across your PDFs, to self-manage workflows end-to-end. Imagine offloading something like KYC, roughly entirely, right, across multiple tools, where Focal AI is able to input hundreds of data points, from conversations, from CRM data From emails and other resources. And then self-navigate the work to be done directly inside of those other tools. So Agentic AI basically abstracts the busy work that you are currently manually doing by yourself. And as for a recommendation engine, today we have the performance coaching feature, that's, you know, baked into the core product, but what's exciting is what can be built on top of that, right? Do you have onboarding manuals? Do you have training manuals? Do you have internal products that you would like to be ensuring are part of that, you know, upsell motion, or part of that relationship with a given client that are opportunistically brought up with that client?

[00:32:51] John L. Connell: Based upon how the conversation is mapping. Some of these others, if we get into sort of how you think about what the technology is, right? Work today saving, you know, advisors that 10 to 15 hours per week. It's, you know, the 30 to 45 minutes in meeting prep, it's the, you know, 30 to 45 minutes after a client meeting when you're trying to update systems tools. If you're an advisor, you're looking at AI as an opportunity, and you're saying, you know, help me save on time, on capacity, help me generate more revenue. And I think one of the big questions that I have, that I'd love to catch from folks in the room here is, you know, if you had that time back, if you had more client capacity, and you had the ability to, you know, pursue, what you'd like with that additional time. Do you prospect clients? Do you deepen client relationships? Do you hit the golf course on Friday, right? So let's actually bring that to the poll as well. If you had more time, right? How would you be saving more What would you be doing with it? And I'll give folks a couple of seconds here to respond to this one while people are coming in on this one, I see one of the questions, we need this at IG Private Wealth, are they in talks of adding Focal AI? If you know Marty or Mike, would love for you to poke them and just let them know that you saw this, and we're actually catching up with them here in the not-too-distant future. I think they are quite excited about the future of Agentic.

[00:34:54] John L. Connell: Super! So I see that the poll questions have come in, in terms of what people would be doing with that additional time. It looks like roughly a third are saying prospecting. Makes sense. The majority are saying deepening relationships with existing top households. Absolutely makes sense. We're finding that the performance coaching capabilities, time and time again, are really unlocking, sort of best-in-class performance. We've seen a lot of folks sort of migrate towards that top 10% producer status as a result of some of the feedback systems there. And then with larger entities, being able to incorporate what products that they're selling, how do you bring more AUM over to your your book of business, and of course, you know, if you had a meeting, say, 2 years ago, that, you know, today suddenly bring up the insurance conversation when you're talking about mortgages now relevant again, being able to find those missed opportunities or oper… or bring up proactively opportunities, is something that we're really quite excited about. And then the third one here that folks are interested in, improving services for next-gen. Absolutely, yeah, inheriting clients. So…

[00:36:03] John L. Connell: What we're finding with new advisors to the firm as well, you know, the traditional onboarding, the training, the ramp cycle, it would take quarters, it would take years before you'd ever feel comfortable with a new advisor in business to get in front of clients, but with performance coaching, you can now get reps in on practicing. And then, it's like having the very best advisor in the industry on your shoulders, giving you feedback on what could have been done better, how do you structure, how do you open up a client conversation, how do you need to be thinking about closing, and the kind of elements that drive success for those relationships that…We're seeing a lot of those new advisors to a business ramp very, very quickly.

[00:36:44] John L. Connell: And in a couple of cases, there's a couple of firms that are actually interested in using, sort of performance coaching, V2, where we're now injecting their data as well, and helping them build out entire training regiments for those new advisors that come to the practice, whether they're net new to the industry or recently acquired, into the practice. Interestingly enough, the last piece here, lowering headcount costs and increasing margin, I'm totally on board with, I think, some of the thinking in the room here as well. Ai is not the Terminator headline that you're hearing time and time again in media. That's a really fun way to position it, right? But the reality is that AI is the Iron Man suit, right? It helps you do more with less. It helps you increase your client capacity. It helps you deepen client relationships. It helps you ultimately, you know, generate more revenue for your business. If you're deploying it effectively. So AI, right, it's no longer the future, it's pretty much table stakes at this point for organizations that are effectively deploying it. And I think we're now ready, yup, to chat on Q&A. So, I'll start taking a look here at some of the Q&A questions that we can go through. If you do have additional questions, feel free to type those in the Q&A section. And David, I see you've also joined the party as well.

[00:38:12] David Kitai: Yeah, I'll be your audience proxy here, John, to put some of these questions to you. But thank you, first of all, for a fascinating presentation. This was a really interesting area, and you can see from the engagement of our audience, I think that there's… this is such a compelling question for so many advisors of what is the right tool to use. I want to start with a question that's come in that's sort of around, you know, the initial tagline of your presentation, the, sorry, what was it exactly? bridge the AI execution gap, or get left behind.

what are the risks if you choose to be left behind? What are the risks to your business as an advisor, or even as a dealer, if you say. Okay, this sounds great, but, you know, we're just gonna focus on what's already been working for us, we don't need the new technology.

[00:38:59] John L. Connell: Yeah, I think this is a fantastic question, and it boils down to a couple of key points. One is that, shadow AI is a massive risk, especially within larger organizations right now. If you're an independent network, right, and you've got dozens, hundreds of advisors that are going off and trying to figure out how to deploy AI, because top-down you haven't been able to manage it just yet, or unlock and authorize tools for the practices. That you're partnered with. What's happening is that advisors are using off-the-shelf tools that are not purpose-built for the industry, and they're having to deal with compliance challenges. So, we've seen scenarios where folks are copying and pasting client information directly into ChatGPT, That's a big no-no. They might be using more general-purpose note-taking solutions that might do a fraction of what's actually possible. And then you've got other scenarios where, beyond the compliance risk, it's really a competitive factor, right? If you've got peers that are now realizing 30-40% uptick to client capacity because they're deploying AI to automate prospecting, to automate onboarding, to, you know, take care of some of the meaning management automations.

[00:40:14] John L. Connell: That is, you know, I think a massive gap in industry today, and that gap compounds as well, as you build out AI agents to service specific points. I think Mark has a really good question here in chat as well, where he says, you know, it sounds like co-work and co-pilot have a lot of very, you know, similar types of capabilities, if you're just working with an out-of-the-box tool. One of the things that I would point to here, if we get into the weeds on Agentic systems is that if you're trying to build these agents on your own, that are with, you know, the power of something like a co-pilot, it's likely, one, going to be missing a lot of the context of the client-advisor relationship. So we have serious depth to the terminology, the 3-4 letter acronyms, the regulatory considerations, to make sure that the outputs are sound. But then, two, in terms of scope of what it can do right? I'm imagining that your Copilot instance doesn't have an AI overlay for something like Conquest, so you can automate, you know, filling out 400 different data points directly within Conquest. I also imagine that, you know, they don't have managed services so that you can layer on prospecting automation, onboarding automation, reporting, getting into custom projects that might be unique to your firm.

[00:41:34] John L. Connell: There is a long tail of workflows that could potentially be partnered on, right? When you think of all the different types of things that advisors, that assistants, that, you know, firms at large are trying to take care of manually today, just to name a few, right? Putting together and generating proposals, bringing in sales analytics data points into some of the training pieces. Yeah, you can set up email and meeting automation across different tools, but is it an all-in-one platform that connects everything, that helps you, you know, identify upsell opportunities, that helps you reconnect with clients and provide those proactive recommendations? So there's a long tail of body of work that can be unlocked when you have an AI partner that is purpose-built for the industry, that can connect everything. and then tie together those workflows, versus something that is, like, a general purpose solution, right, that is more of an off-the-shelf relative to built for advisors and for large enterprise practices.

[00:42:41] David Kitai: Fascinating. So, there's another question that's come in here, and I may be doing some interpretation here, but the question, as it's worded, is what's stopping a client from using AI to do this themselves? And maybe the way I'd interpret it is why can't I vibe code my own version of Focal? Why use a vendor like yourself who's built a platform, especially in the age of, you know, clogged code and all that kind of stuff?

[00:43:04] John L. Connell: Yeah, yeah, the biggest challenge there on the Vibe coding piece is likely going to be around compliance, right? Are you dealing with publicly exposed URLs? We've even seen, like, notetaker tools that are, in fact, SOC 2 Type 2 compliant. That have publicly exposed URLs. That is not compliant with our industry. Are you abiding by implied consent under PIPEDA? For the folks that are following a lot of the guardrails frameworks that come out of the OPC? How are you thinking about provincial regulation considerations here, right? There's a lot of gray area, but there's also a lot of guidance that's already out in the wild today. So if you're not operating, you know, at a very conservative level, you could be looking at the risk of potential fines in some cases. And there's a lot of ways that this turns into failure mode, right? If you… let's say… one of the things we've come across, folks that are using general-purpose AI, they'll say, hey, you know, I've got my MD file, or I've got my prompt, and I've been able to sort of generate notes for a very specific types of meeting conversation.

[00:44:07] John L. Connell: Which is great, but then I ask, well, how do you tie that context in with your CRM, with your planning tool? Do you need to generate proposals? Do you need to, you know, automate your prospecting workflows? Do you need to look into connecting with bespoke Systems tools? Do you have retirement documents on, you know, PDFs that effectively could be filled out on behalf of you? You look at opening up a new account, for example, as part the onboarding process, right? That's a pain in the tail. Now, what if AI could come in, help automate all those steps the filling of data, the filling of PDFs, you know, emails that get authorized and sent off. With the enterprise relationships, that's where we can really get in the weeds, and help organizations unlock the value of Agentic AI, because it's great that we can come in and save you, you know, 10 hours a week with the core asset today, the meeting automation, the performance coaching, the data entry automation. But the other, like, 80%, 90% of value with AI actually comes from Agentic automations. It comes from deeply understanding the personalized workflows that, you know, are the clicks across different systems and tools. If you have a single pane of glass, right, that basically connects all the systems, all the tools, and can take action across the entirety of your back office on your behalf.

[00:45:26] John L. Connell: And it's doing that compliantly on guardrails with human in the Loop, right? That is a very, very different approach than trying to, like, independently build out these, you know, general purpose capabilities, or going after point solution capabilities. There's a compounding effect to deploying Agentic. Once you have that data, once you have some of the core platforming, the value that can be unlocked that sits on top of that is quite tremendous, but…I think a lot of organizations right now are trying to understand, the square one on, you know, is this thing compliant, and are we using the right decision-making matrix to find the right vendor to be deploying with? And that answer's going to be very different from firm to firm, right? So what we've done is we've said at Focal, hey, you can come in and you can unlock immediate wins with the core platform, right? But then if you're interested in having an AI partner We can bring those services, those capabilities to your practices as well, to help you really drive that Agentic automation.

[00:46:27] David Kitai: Another, another great question that's come in, just…Where do most Canadian wealth firms kind of slip up on PIPEDA and provincial compliance when they're deploying AI, especially around something that you mentioned earlier in our talk about vibe coding, implied consent, thoughtless meetings, audio-video storage, and using customer data for LLM training? And what questions should executives be asking every AI vendor in their due diligence?

[00:47:53] John L. Connell: Yeah, yeah, this is a really good question here. When you think about the compliance pieces, some of the big considerations are here on this slide. So when you're thinking about the data residency, where's the audio stored, the video stored, is it stored, right? Are you using stateless AI architecture? You don't want your client data to be trained on. That's a big no-no. Soc 2 Type 2 is a must. One of the big misses that we see, in a couple of other vendors, actually, in our space, interestingly enough, is around implied consent with PIPEDA. So we take a very conservative approach with implied consent. We make sure that, you know, we're not taking a bot-less approach. Which, if you're starting to transcribe conversations before you've had consent granted from your clients, right, that can create regulatory exposure and potential fines to be looking forward to.

[00:47:50] John L. Connell: So what we really make sure to have visible and auditorily engaged is the sort of intention of the bot being very clear going into that meeting. The, the other pieces around compliance, if anybody's interested in going down the rabbit hole on specifics there, please do reach out, because this is one of those topics that, if you are not set up…architecturally sound under the hood, and you are not abiding by implied consent, or, you know, some of the core requirements, the guidelines, as well, that are outlined around communication technology. you can find yourself in a really, really tough spot, so…like, if you're using a general-purpose note-taker, right? A lot of those have publicly exposed URLs. If you're using something that doesn't, This is a long conversation. But I would say, please make sure that whoever your vendor is, that they at least check the box on these four big pieces here. If you're a CCO, there's certainly more to it than this, though, that we could chat on.

[00:48:58] David Kitai: There's potential for a podcast for an audience of about 25 compliance officers, that we could run, and they would be glued to… to the discussion, I have no doubt. Because it is… it is utterly fascinating what's… what's gonna… what these… finding the answers to these questions now. Going beyond that, though, looking at it from almost a dealer perspective, we've had this question come in, which is, beyond our saved per advisor, which was what you've kind of ended with. What KPIs and leading indicators should executives track on how an AI deployment is actually working? And at what point do you decide to scale a pilot firm-wide versus hold the plug? Getting into the joys of the sun cost fallacy here.

[00:49:43] John L. Connell: Love it. Yeah, this is a great question. When you think about outlining ROI, and then making sure that you're actioning on the ROI that you're realizing as a business. From a metric standpoint, it will depend upon the problem statement that you're trying to tackle. In our case, if we're looking at client meeting management automation end-to-end, right, there are a number of sort of short metrics and then high-level, star metrics, that we could try to qualify, so…Looking at the time savings is certainly a big piece to this, right? The other big problem statement is around the elevating of performance for advisors, and so what does that mean, and how do you measure both the time savings and also the performance? On the time savings side, if we break down you know, the work before, during, and after meetings. Today, it's, you know, the 30 minutes that you're saving from the agenda Because we're generating that in a matter of seconds. It's the email that you previously had to take time to craft, that we're now generating in seconds excuse me, it's that client overview, before the conversation that we're also generating in seconds. And then during the meeting, instead of having to meticulously take notes, right, this is the harder one to measure here, is the value of being able to focus more intently on your client.

[00:51:02] John L. Connell: Being able to double-click on specific questions because you're not trying to just jot down as much as you possibly can and, you know, miss potential details. And then the, sort of, post-work around the, you know, organization of those notes, structuring the tasks, pulling those tasks out, piping this data into your CRM, piping this data into, you know, planning tools, that's 400 fields in Conquest that we're able to autofill, right? That's where a lot of, sort of, the metrics start to become pretty, you know, crystal clear, and you can break that down by advisor, too. So what we do with a lot of our pilots is we'll help organizations understand, here's how it is that you're saving time before, during, after work, like, literally breaking down the hours saved every week.

[00:51:49] John L. Connell: And then here's, like, the total time that your organization's saving. And you can map that to dollars as well. So you can say, okay, based upon the 200 hours, you know, that we've saved for your organization this week, that equates to roughly X number of dollars. And if we annualize that over a year, you know, you've now got an uptick of another 17 or 18 advisors in terms of client capacity that you could be unlocking for your business. Right, so you can talk about it from the advisor dollar standpoint, you can talk about it from client capacity. The other side of the coin here when it comes to performance, this one is a little more fudgy, right? In terms of how you think about behavioral finance and how advisors improve.

[00:52:31] John L. Connell: The numerical, though, is around better prospecting and better AUM conversions. So, if you can bring over more assets to your business as an accelerated rate, or if you can prove out that you're closing more prospects because of some of the behavioral coaching, and then when we get into injecting, you know, additional products and onboarding and training into your personalized instance of Focal for your firm right? There are, again, dollar values that can be associated with that, and you can see pretty quickly, you know, bottom quartile advisors turning into top 10% producers, and you're like, well, what caused that? Right? It's gotta be combination of the performance coaching, it's gotta be, you know, other variables from the time savings. Maybe it's going deeper with existing clients because they have additional time, so they can, you know, offer up more services and, you know, take somebody out to lunch on Friday, since they're saving 8 hours, you know, in that given week That they otherwise wouldn't have been able to have done.

[00:53:33] John L. Connell: So it's… it really comes down to…those two big variables, though, it's like the time back, right? How does that equate to dollars? And then the improvement in performance. If all of your advisors are suddenly operating at their best, and you have visibility on how your best in your practice are performing and what they're doing differently, and by the way, you can now bring those learnings to everyone else. That is very powerful. Not to mention the ramp time. We don't even capture this today, like, of new advisors or new practices to a larger entity. You know, how they're more quickly ramping or sort of acclimating to that new environment, that's a whole…You know, sphere of just benefit that we're not even looking at right now.

[00:54:16] David Kitai: No, fascinating, and it touches on so many things, the sort of more time to be human, but also that kind of poll you had earlier, which where nobody said they would use AI to necessarily look at headcount or margin, it seems like it's accretive to margin, even if that's not the core goal. It's that by adding these productivity gains, there's an improvement in the bottom line, which is a really interesting kind of way to look at it.

[00:54:41] John L. Connell: And the other piece to this, David, is that that improvement compounds dramatically, right? Once you have an AI layer in place, all of the other things that you can do to automate, to offload manual work, you could effectively… anything that is human clicks in a system navigating different systems and tools, right? You shouldn't have to go to 4 or 5 different tools for an address change or a beneficiary update, right? Instead, with Focal AI, you click the go button after your client conversation. And then that seamlessly goes across all the systems, all the tools, fills out the paperwork necessary, and then you've effectively offloaded the entire process of what today is very manual re-keying into different systems, tools, and paperwork, right? And that, of course, is just one small example, right? When you get into the compounding nature of deploying AI, if you can think it, meaning something that humans type or click.

[00:55:36] John L. Connell: Right? AI today, with intelligent robotic process automation, can effectively do it. So, one of our big goals at Focal AI is ensuring that advisors are able to go as deep as they can with their clients, focus in fully on those relationships, and then, as your AI partner, we're cherry-picking and offloading more and more of the manual, just, toil that you have to deal with day in and day out. And that's why, you know, this is oftentimes as much a play at the center of the enterprise, or the broker-dealer network as it is at the independent firm, right? And so, the organizations that will most effectively unlock the value of AI, and frankly will be winning in the long term, are those that are building in the center, enabling these capabilities across their entire network.

[00:56:25] John L. Connell: Instead of the reverse, where we've seen, again, the failure mode of the marketplace model on, you know, advisors go figure it out, we'll just authorize a bunch of tools, then you have to rebuild the wheel with 7, 8 different vendors, and of course the vendors are then not incentivized to like, build Agentic AI for you or your organization, because they have a small handful of firms, right, to be deploying with. So, the orgs that are partnering, going deep, gathering that data context, and then unlocking Agentic for their entire network. If I'm an advisor, I want to be in those networks that are, really thinking intentionally around operating AI at core, and then unlocking those capabilities for my practice, and importantly, going fast, right? If you're at a firm that is trying to build this all out internally, we've seen that happen a couple of times, you're probably gonna you're going to be going a little slower than something like a firm that's partnering with a focal AI. That can help you go much, much deeper and much quicker, because we've already built out a lot of the types of automations that you might want to be deploying for your network of advisors.

[00:57:32] David Kitai: Okay, that is fascinating, and I think a really excellent summary of kind of the calculus for a lot of firms and advisors to end on, John, as we come very close to the end of our session. So, with that, all I really have left to say is thank you so much for this presentation, thank you for a really fascinating discussion, and for Touching on, you know, what clearly we can see is such an important topic for the industry. And finally, just a note of thanks to all of our attendees. Really appreciate you taking the time out of your busy days, appreciate the work that you do. For WP, I've been Dave Kitai. Have a great rest of your day.

[00:58:10] John L. Connell: Thanks, everyone, for joining. Appreciate it, David. You have a good one.

[00:58:12] David Kitai: Be well. Bye, everybody.