Glenn Elliott, co-founder of Practifi and founder of Meldus, explores how artificial intelligence, automation, and open platforms are reshaping the financial advice profession.
—
In this episode of The Trusted Adviser, Rob Pyne sits down with Glenn Elliott, co-founder of Practifi and founder of Meldus, to explore how artificial intelligence, automation, and open platforms are reshaping the financial advice profession.
Glenn shares insights from more than a decade leading Practifi’s global expansion and now from the cutting edge of AI-driven tools at Meldus. From CRM evolution and the power of open APIs, to practical use cases of agentic AI in advice firms, this conversation uncovers where the biggest opportunities lie for advice practices wanting to scale with purpose and efficiency.
Whether you’re a financial planner, practice owner, or industry professional curious about the next frontier in technology, you’ll come away with clear ideas on how to future-proof your firm.
LISTEN
‘
SHOW NOTES
Topics discussed
- Glenn’s journey: from Practifi co-founder to AI innovator with Meldus.
- The original inspiration behind Practifi and its mission to create a single source of truth for advice firms.
- How AI and automation are transforming workflows in financial planning practices.
- Why open platforms and APIs are critical for firms looking to leverage technology.
- The difference between RPA (Robotic Process Automation) and API integrations – and when to use each.
- Real-world applications of agentic AI for client meetings, compliance, and practice management.
- The shift from manual workflows to AI-powered decision-making and quality control.
- The importance of data quality, structure, and access in unlocking AI’s potential.
- How Microsoft, Salesforce, and OpenAI are shaping the AI ecosystem for advice.
- Glenn’s personal use of tools like ChatGPT (voice mode) and his reflections on how voice AI is changing the way we think.
- Advice for growing advice firms on building an AI-ready, open-architecture tech stack.
Quotes
- “The AI opportunity is there to be seized, but it starts with you having access and control of your own data.” – Glenn Elliott
- “Agentic AI takes us a step beyond workflows. It can analyze across your entire book of business, prioritize what matters, and tailor insights to your role.” – Glenn Elliott
- “If your platform isn’t open today, that’s a red flag. It should have been open 10 years ago.” – Glenn Elliott
- “I think the future of advice is bright as hell. AI isn’t replacing advisers—it’s giving them more capacity to be human, trusted, and valuable.” – Glenn Elliott
About Glenn Elliott
Glenn Elliott is the co-founder of Practifi, a CRM purpose-built for financial advice firms, and currently serves on its board. In 2023, he launched Meldus, a technology venture focused on applying AI and automation to unlock CRM data and create more intelligent, user-friendly adviser experiences. Glenn is recognized globally for his contributions to advice technology, blending visionary design with practical application.
Connect with Glenn on LinkedIn.
Resources & Mentions
- Practifi – CRM for advice businesses: practifi.com
- Meldus – AI-driven insights for advisory firms: [meldus.com] (if applicable)
- References to Microsoft Copilot, Salesforce Einstein, and OpenAI’s GPT models.
Key takeaways
- AI is your new tech teammate – Choose platforms AI models know well to boost efficiency and smarter decision-making.
- Leverage AI’s knowledge base – Open platforms with well-documented APIs give AI deeper insights and stronger support.
- ChatGPT as a research partner – Use AI to compare platforms, uncover limitations, and make better-informed tech choices.
- Faster prototyping than ever – Test, iterate, and refine ideas quickly within secure, robust environments.
- Stay nimble and adaptive – The rapid pace of AI innovation requires flexibility to stay competitive.
- Expand your technical capacity – AI extends the capabilities of your team by tapping into vast data and documentation.
- Better vendor choices – Select platforms AI understands deeply to ensure smoother integrations and reduced risks.
- Mindset shift required – AI isn’t just a tool; it’s an extension of your technical team, changing how you approach problem-solving.
TRANSCRIPT
Rob Pyne
Welcome to The Trusted Adviser podcast, where you get a deep dive into the world of financial planning with industry leaders who share their stories of winning and learning as they chart their path to success. This podcast is for the curious. Those of you who like to dig into the detail, and if that sounds like you, get ready to listen and learn, and if you’ve been here a while and you’re getting value from these conversations, I’d really appreciate you subscribing or leaving a review. It helps others discover the podcast and join the conversation. Today, I’m joined by Glenn Elliottt, co founder of Practifi the CRM platform, purpose built for financial advice firms. Glenn helped launch Practifi back in 2013 leading the company through global expansion across the US and Asia Pacific, before stepping into a board role last year. Over that decade, he worked closely with advice businesses to understand their biggest pain points and build solutions that improve productivity, compliance and client engagement. Since transitioning from the CEO role, Glenn has continued shaping the advice technology space, most recently through Meldus. His new venture focused on AI driven tools that unlock the value of CRM data. With his unique perspective as both a tech founder and board advisor, he’s in a prime position to talk about how AI and automation are transforming the advice profession where the biggest opportunities lie, and what firms need to do now to future proof their technology strategies. Welcome Glenn Elliott to The Trusted Adviser podcast. Thank you, Rob. Great to be here. Really good to see you. Great to see you again, Glenn, we’ve known each other for a while now. We were Practifi users back in the early days, you’d started that in 2013 and we signed on in 2017 but before we go back to your history and the journey you’ve been on in technology as a co founder and technology entrepreneur, I want to take a second to reference this comment you made on LinkedIn. Last week, you saw an article on money management saying, Will solo AFSLs go the way of the dodo. And you made a comment there and said that you’re inspired to show independent financial advisors what’s possible with a creative, nimble approach to AI. And if you’re self licensed to wondering how far you can push the boundaries efficiency, let’s talk. So that’s what we’re gonna do today. We’re gonna talk about AI automation and what’s possible, and what you’re seeing in your vast experience in the technology space. So we’ll go back to your Practifi journey. 2013 is when you kick things off with Adrian Johnson, so your co founder, what inspired the two of you at the time to create a CRM platform tailored specifically for financial advice firms?
Glenn Elliott
Well, look, it was a fascinating journey, and it started from our consulting work in the wealth management space. We found ourselves doing the same kinds of engagements, the same kind of project roll outs over and over again. And there was a real trend that emerged. We would work with advisory firms who struggled to get access to their data, who had silos of data, client information in this silo, and this database and this platform usually fragmented and disconnected and just crying out for reorganization and integration and just an open approach that puts the client at the heart of the whole business. That was the kernel of the idea for practify. We found ourselves designing and implementing those solutions repeatedly. People don’t want to buy this on a custom project basis. They want to buy it on a software as a service basis. From that Practifi was born.
Rob Pyne
so during your decade at the helm of practify, what were the key pain points in financial planning businesses you were aiming to solve that bringing together of the data sets that are all fragmented, trying to sort of coordinate everything in a more cohesive fashion. But how did the platform evolve? And now that you serve on Practifi board, what do you see as the next frontier for CRMs in the advice profession, especially in the context of AI and automation and what that’s bringing to the table? Look,
Glenn Elliott
our job with Practifi was always to create a better experience for advisors, for their teams, and ultimately, for their end investor clients. We wanted to smooth every aspect of Operation across the business that we possibly could. We always had this role based approach to our system design, so that the role of the advisor was distinct to that of a client service representative to a compliance manager and others in the firm would each have their unique take on what was an integrated single source of truth about their relationship with their clients, and just streamline and make more effective and more automated every step in that journey that we possibly could my great passion as a tech founder is always to create something that is elegant, that’s beautifully designed, that is fit for purpose. It’s our job as tech designers to think several steps ahead. It’s one thing to listen to your clients, listen to the market, and hear their pain points and understand. They want today, but there’s a, you know, there’s an art and a science to software design, and it’s our job as technologists to think ahead of that and say, Yes, I I know that you want that process streamlined, but let’s question the process in the first place and determine how does that fit into your overall set of processes. And can we look several steps ahead and maybe get you further ahead in one leak, rather than a lot of incremental steps. So I always had that philosophy at Practifi, and that continues today, even from a board level, as the company emerges quickly into AI automation, but still with that same heart of a fully integrated source of truth that’s really meant to make the lives of everyone in that advisory firm better.
Rob Pyne
Yeah, as I said in the introduction, you stepped back from the CEO role into a board role at Practifi, really, to take on new opportunities, new ventures, because you’re clearly someone who’s always got a myriad of ideas that you want to try and explore and implement. So you launched Meldus After leaving full time operations in Practifi, although you’re still on the board there, what insights from your decade at Practifi have directly shaped the vision you have for your new venture. Meldus.
Glenn Elliott
There was a really kind of clear lineage there. So two fold. We always had a cohort of users at Practifi who I would put my heart and soul into the design of the software and every aspect of the user experience. Me, my team, my team gave me a terrible reputation. They used to call me the pixel Nazi, but it didn’t matter how they laid out the screen. I would look from a distance across the end of the room say, aren’t that that needs nudging there, that needs and it was never quite perfect enough for me. We cared deeply about that user experience, that ease of adoption, and how close we could get people to a zero training experience is always the aspiration, and the advent of generative AI takes that really one step further and allows you to create user experiences that are just even smoother than that. That can be just natural language activated, can be voice activated, or can be auto activated, so that you get insights and information from your data before you even ask that was sort of the Holy Grail. These were mythical concepts back in 2016, 1718, as we emerged in the market and expanded to the US. But now in my move to Meldus, all of a sudden, they’re possible. So we’ve created this kind of agentic framework and applied this in various ways that just smooth that experience for totally non technical users. And I find that really exciting.
Rob Pyne
Yeah, you speak about elegance and design, it sounds like you’re very inspired by the Steve Jobs approach to technology. Everything was about design and making it the great user experience, whether it’s a physical object or whether a Software User Experience is everything, the way that he’s talked about, I guess, in the context of what was Steve Jobs, great skills, communication, but also his ability to kind of focus on design and be a bit of a whether it’s pixel Nazi, as you describe yourself, but getting design so that it makes the user experience so great that people want to use the product, and in your case, your software as a service, which Practifi. So can we take a step into the sort of AI world as you see it? Now, you know from your experience with advice firms, how is AI transforming the way planners use CRM data, and where do you see the most realistic productivity gains coming from for financial planning firms?
Glenn Elliott
Well, starting with the CRM is a good place to start. That continues to be a varied playing field in the industry, and particularly in Australia. I’ve spoken to a number of advisors recently who are excited about the advent of AI and what that means for their firm and for them and for streamlining their operations, but getting access to their data in CRMs and other platforms can still sometimes be stuck behind legacy APIs that are proprietary, not well documented, not well known in the developer community, sometimes they don’t even exist at all. And so even just as a fundamental foundation, it is more important than ever before for advisory firms to be working on open platforms and to really drive their vendors to an openness of information architecture, no less secure. Of course, these are two sides of the same coin, but a commitment to open standards and open access to your data in a very controlled way, that data is yours. The AI opportunity is there to be seized, but it starts with you having access and control of your own data, but once you have that foundation in place, the opportunities are manifest. Now there are so many firms talking about what is possible in future. What is the future of AI and what does this mean for the future of work of the industry? Do we need advisors anymore? I’m really not fixated on those questions. I think the future of advisors and their businesses and even smaller AFSL like you point to in that money management article. I think it’s bright as hell. I really do you have the opportunity now to be an even more trusted and human advisor with an even stronger relationship with potentially a larger number of clients by streamlining and automating. The mundane and the everyday, the stuff that you clients don’t care about, they want to see more of you and get more advice from you and leverage more value from you. And I think that opportunity now is greater than ever if you get control of that whole information architecture and landscape in your own firm.
Rob Pyne
Yeah, not every technology vendor is open, though, are they or they’re resistant to sharing their data source, or, you know, opening themselves up to an API and connecting to other software platforms. So you see that as a real impediment to businesses that are using them, if they do continue to use them. And what do you see as the eventual outcome if everyone kind of comes to this same realization that you have, that you have to partner with firms that are open, that can connect vendors, that is technology vendors. Where do you see the future of those firms that choose not to do that? And we know kind of the incumbent in Australia is moving away from they sort of were very close. They went to a more open stance, but they’ve seem to be retreating from that now and reverting more to a closed shop type approach to managing their business and their data source. How do you see that playing out? Glenn, from your experience and what you’re seeing how AI can transform businesses and productivity gains. What do you see just roll forward five years? What does the landscape look like, given everyone’s kind of really interested in where this is going to take businesses, how AI and automation will enable business to be more productive. And I was talking to a practitioner last night who said they’ve moved from a two to one ratio of support staff to advisor. So two support to one advisor. Their business has been investing heavily in technology, and they’re nearly one to one now because a lot of the back office support systems and people have actually been automated, and that doesn’t mean they’ve actually dropped anyone in the process. They’ve actually simply not hired new people as they’ve gained more advisor staff, they’ve simply not added more support staff to back them up. So it’s a long question, but where do you see the future being? Let’s say, in five years from now, with AI and automation and businesses as you know them, and the landscape of technology vendors in that context. Let
Glenn Elliott
me kind of answer in two parts and perhaps in reverse order. I love that story about it improving the ratio of support staff to advisors, but I hear exactly the same in that firms have worked so hard to recruit and retain tremendous talent in their organizations, and particularly those of a technophile nature, they’re hard individuals to find, and once advice firms have them, they tend not to want to let them go. And so what I keep hearing is the same as you this is an opportunity to make the day to day lives of those staff even more enriching and more purposeful by streamlining and automating the menial the ratio that I keep hearing coming through is, how do we increase the ratio of clients per advisor and revenue per employee to retain the great staff that we have, but grow the business from that same input base? And I think that’s a huge opportunity to your question around vendors. It should be a red flag now, not in three years time. Five years time, if your major platforms, you or even your minor ones that you’re relying on in your advice for, particularly those that are the sources of truth, they should be open now. They should have been open 10 years ago. We had this philosophy at the start of our Practifi journey, and that’s a long time ago. Now, there are firms doing clever things with AI, using RPA, using robotic process automation, so going in through the user experience with AI, for AI to effectively operate software as a human being. And that’s clever, and it really has its place, and it has its place for legacy tools that don’t have a system level API for your AI to connect to, because which is always the better way. It’s more controlled. It has clear security protocols. It has clearer audit trails. That’s the preferred model. So if you find yourself dealing with closed platforms where your only AI oriented approach is to go in and pretend to be another user. That’s a red flag, and it should be the worst case scenario. The industry has adopted Microsoft, 365, for example, very, very heavily across the space. It’s almost something of a standard, and one of its great strengths is the openness of the platform and the ease of working with APIs. And there’s millions of developers around the world that know how to do that. That’s the sort of position you want to get yourself into. I mean, we built Practifi, I on the Salesforce platform for very much the same sort of reasons, open APIs, accepted standards that technicians know, which allows you the flexibility in advisory practice to have this huge marketplace of really strong technicians that can work with your information in a really controlled and secure and industry standard manner. If you don’t have that available to you, that should be a red flag right now.
Rob Pyne
Yeah, can I just pick up on a point there you made? Glenn, you said, robotic process automation RPA as it currently functions. In fact, I had a great chat with Nick Perrett from Yarra Lane Group, who’s doing exactly that last episode. Yeah, so fascinating stuff. But you also said there, a moment ago, that there’s also another approach. There sort of more of an API approach. Can you just contrast those two things for me, because I was super interested in what Nick was doing with his RPA, but also realizing that as long as that process is very standardized and not likely to change or shift greatly, then you can employ that, whether it’s review, preparation for client meetings coming up if a standardized approach, I think Nick uses it heavily in their tax practice, pulling data from ATO portals. They’re doing it in their advice business, pulling data from BT panorama. So can you just contrast you said there? I think, and maybe I got this wrong, but there’s an also an alternative approach there, using an API approach which is more flexible than the sort of RPA approach, yeah. So
Glenn Elliott
let’s use a Microsoft example. So we’re all familiar with email and familiar with Outlook. Most of us use it. Obviously, you use your email through Outlook, and you get your inbox there. You can open up a given email and you can read it. Read its contents using an RPA robotic process automation approach. You can use an AI agent or AI bot to simulate your activity as a user, to open up Outlook, to go to your email list, open up one and deal with its contents. And that’s fine, but it’s really not the vendors preferred approach, Microsoft and all other open technology platforms that lead the way in this. That’s not really what they want you to do. They want you to use their application programming interface, which is, it’s just another lens on the same data. So instead of the user’s lens, it’s a system to system lens, and they’ve gone to great lengths over decades now to perfect those APIs make them very, very standard and robust and high performance and scalable and all those great things and secure, obviously, so that your AI agent, in this case, can actually communicate through the API to access, Say, your email list, and access one email and get it its contents the same way. That’s more robust. It’s vastly better documented, and that is by far the preferred approach for sort of seasoned technologists in the marketplace. The RPA approach is very handy when you don’t have APIs as an option, and that will be the case on an ATO website or an ASIC website, or perhaps some investment platforms and the like. But it really should be the fallback. If a standardized strategic API is not available, it gets a bit nerdy, but it’s exactly those sorts of decisions that once you put them behind you in your advisory firm, then you realize, okay, great. Okay, well, here we have a really strategic approach to that integration. This one here’s where we need a fallback, and we’ve made very clear decisions at every step in that way.
Rob Pyne
Yeah, and that’s really good insight. I’m glad I asked that question. Follow up. Just your comment there, because that helps me understand that whole landscape. And I do understand the API approach, but I didn’t quite make the contrast before, so that’s really helpful. What parts of the process do you anticipate will be the most impacted by AI driven tools? You know, like those you create at Meldus. In the near term, we’ll talk more of Meldus and what you’re doing there, which parts will continue to rely on human judgment, and which parts will be really impacted by AI driven tools? Do you think
Glenn Elliott
we’re seeing this kind of emerge in waves? And so some of the first cabs off the rank, if you like, firms looking to automate the mundane and automate the menial, repetitive processes that they do today. And this makes perfect sense. You mentioned earlier, getting meeting transcriptions and automating file note generation from those meeting transcriptions, perhaps generating follow up task lists from those transcriptions is just very worthwhile and powerful stuff planning and automating the process ahead of a client meeting or an annual review, something like that, for reviewing and updating your APL, automating that research of products available in the market is very, very powerful, and firms are doing it today, the next sort of phase of opportunity beyond that, and this is where we really tend to help, from a creativity perspective, is to enlighten firms to more of what’s really possible. Some of this is to think of AI, not just in the context of your relationship with one client, but across, potentially the whole business. If AI can give you insights and automation around one client file, what can it do with broader aggregated ability across a book of clients, or about all the clients you need to meet next week, or about all of those who have an annual review coming up next quarter? This is where this agentic AI concept comes in. This is really almost sort of 2025, concept. We couldn’t do this 12 months ago, but now we can the window of information that you can make available to these AI agents now is 10 fold what it was last year. And it can digest that information. It can read i. Suggested it can reassess, it can validate its own original thinking and update its own thinking with multiple iterations of the same information. And that creates a new realm of analytic power that humans haven’t done before. We haven’t been able to do this as human beings at that scale, and so I love to just kind of feed the information for advisors around, okay, what sort of insights could we get that would streamline your planning for the next three months based on all of the information across your entire book of business? And that gets really a next level of visionary
Rob Pyne
so just paint a picture for us then about the vision of what’s possible there. Just give us some things to kind of like aspire to, or some really visionary ideas that you’re thinking about, thinking this is what I could see being possible as a result of what we’ve now can do with a broader data set being available to agentic AI, looking at your data in more complete way than you’ve ever been able to do, or ever could do as a human being. Do you have some really interesting and visionary ideas about where you think that’ll take advice businesses,
Glenn Elliott
yeah, and I’m hearing more and more of these all the time, and they all come from this perspective of top down aggregation, the ability to sort of wake up on a Monday morning and have a sense of what’s critical for you to consider in the near term, and perhaps in the medium term thereafter, this could be critical dates coming up that are meaningful to your clients, expiring identity documents. So your critical family dates you know kids, university graduations, or you know forthcoming retirement dates and the like, to be able to assess on mass the impact of a proposed regulatory change or a forthcoming possible change in superannuation legislation before it hits to say, Okay, do a what if scenario. Who would this impact across these 100 client files, and who should I prioritize first, if this was to come to be, who should I have the first conversations with, who’s impacted most that sort of aggregate level analysis? It can be done two fold. It can be done in conversation with an AI agent, and that’s where these things have started. That’s where we started with Meldus, was to build this conversational interface into complex data. But it can also be done autonomously and in a really personalized way. As you carry on that conversation with the agent, it can come to understand what you care about, what you care less about, and build up a personalized profile of you and your role and your remit within the business, your targets and ambitions for the year ahead, perhaps, and it can tailor its insights, both through conversation and autonomously, as well as to what really matters to you. This move towards agentic AI that personalization control combined with the autonomous, sort of multi threaded nature of it. It’s such a big step ahead of where we were even 12 months ago. 12 months ago is much more AI oriented workflows. Now it’s aI powered agentic decision making, and the ceiling on that has really just been lifted dramatically.
Rob Pyne
Yeah, I’ve heard similar thoughts shared that you’ve just described, and it doesn’t seem like it’s a long way away now, like that prospect of what you’re describing there seems like a pace at which things are accelerating around AI, and what it can do is it realistic to think that sort of capability is available today, or in a couple of years today, from today, in terms of what people will be able to you see implementing their businesses as a routine thing, as opposed to being sort of the just the tech leaders that are early adopters.
Glenn Elliott
Some of this is absolutely available today. The starting point is access to data. And so in the first AI agent that we created at Meldus, the first data source we used was a connection to Salesforce, and this is any flavor of Salesforce. This works well with and it started last year as an AI question and answer type of Assistant, where you could ask it detailed questions of anything across your client information on that platform, and it would come back with really useful answers. The second iteration of that we created this year is this agentic approach, and there’s a really standard design pattern that’s starting to emerge in the industry here, this idea of creating an agent and giving it tools. And so you might give it a tool that can access data on, say, Salesforce platform. You give it another tool that can access data from your files held in OneDrive within Microsoft. 365, another one that might be able to access your email and calendar through the API that Microsoft provides. And all of these are secure, fully authenticated industry standard type connections that you’re totally in control of. Another might be access to your policies and procedures, your AFSL, your own guidelines about how your firm is run and what matters to you. And so then you start to build up capabilities of an agent that can work with you and understand what you need from your data and can determine. In how to aggregate and synthesize the data across those sources, but using its own decision making power, it’s a real shift, and so that’s quite feasible now, and we do this already today. But where that’s going is to have multiple agents that work in harmony. You can have that sort of agent that I just described. What we’ve started doing recently is creating other agents to check the work of the first agent. So these sort of QA, these quality control agents, are really, really powerful. This is great from a compliance perspective, to go and review a whole swathe of client files and make sure that they adhere to your standards and your policies and procedures. It doesn’t replace the processes and relationships that you have today and supervision responsibilities you have today, but you can really streamline and add a layer of automation in there so you can be more operating by exception than trawling through the detail.
Rob Pyne
Yeah, I love that. What prerequisites around CRM data, quality and structure are essential for advice firms in order to unlock these benefits of AI tools, I mean, there’s clearly going to be a way of capturing your data and storing it such that AI can be most effective with your data set. So can you tell us a bit about that data quality and structure that’s necessary to make it effective?
Glenn Elliott
So it’s first and foremost the openness of the platform. So most platforms in the industry, whether they’re more CRM focused or whether they’re more portfolio management focused or advice execution focused, most of them have quite a thought through structure for their data, particularly the mature tools, and most of those are quite good. So when you create an agent like this, you tune it, you say, right, here’s access to my data through an API, preferably. And here’s some detailed documentation on how that API works. So there’s a client’s table and there’s a securities table, and there’s a transactions table, whatever they may be, there’s an interactions table, and it can ingest those instructions and learn from those instructions and access that API make sense of it with well thought through data structures like that, which the industry tends to have, but they don’t all have the openness. It’s getting that access in the first place, which is so critical. So if you have that look, the more standard, the better, because, of course, these language models today have an inherent understanding of some platforms. They understand the Microsoft 365 platform intimately. They understand the Salesforce platform and its various derivatives intimately. They understand esoteric industry tools far less well, but it’s still possible. It just takes a little bit more tuning on those data models and information architecture, as long as they’re open.
Rob Pyne
Yeah. Okay, so you’ve talked about Salesforce and Microsoft and both have, obviously, market leading CRMs, and they both have an element of AI built in copilot within the Microsoft environment. And Einstein is Salesforce is equivalent, I guess, and you know it better than me. But how do you see businesses looking to adopt AI in their tech stack? Do you encourage them to go more than native? Built in tools that are being built by these CRM leaders, or are bolt on tools just as effective if it’s an open architecture platform CRM the way the Salesforce and office 365 are.
Glenn Elliott
Well, the great news is that it needn’t be either, or it can be a blend of all of these. Something comes down to commercials, as much as it does technical decisions as well. Some of the commercials around those built in tools have been evolving very quickly, and generally they’re coming down. The prices are generally coming down. Microsoft copilot launched as an add on price and then started to get absorbed into certain versions of Microsoft, 365, which is very promising, and I know many advisors are making pretty good use of that in their firms. Salesforce gone very hard at all things, agent, force. So every brand message this now sings this from the rooftops. It started as slightly alarmingly expensive, but is now getting more affordable. It’s a complex pricing model that they have. But if you’re on that platform, it’s it’s worth exploring, but I would absolutely argue that those sort of major incumbent vendors in the world of AI are just not moving at the same pace as others. And I applaud both of those platforms for having such open APIs to allow their customers to work with third parties in the industry, Microsoft has done a great job with their association with open AI, for example. So one of the great opportunities in wealth management is to have that backing and enterprise security commitment that you get from Microsoft, but to have access to the really leading edge open AI models like GPT five, for example. And there are many, many other AI models around, but that’s a very safe bet that just gives you that confidence of the compliance standards that you need with the commitment to open APIs are really leading edge models, so there’s a lot of flexibility, even through that stack alone. And this is my point in the money management art. Cool that’s available on a transactional level, cost that’s the same for one staff member in a firm as it is for 1000 Yeah,
Rob Pyne
yeah. It’s incredibly encouraging, isn’t it, that prices are being pulled down just by market competition. I mean, it seems to me, as a keen observer of what’s going on, that open AI’s approach to really pushing it out there into the market, chatgpt and its delivery of new models at a pretty rapid clip. It’s forcing others to take a more aggressive approach as well and not be left behind, because OpenAI really trying to sort of steal the entire market. And they kind of had a really early advantage there by being first to market. But just on a side note, what models are you using routinely in your daily life, whether it be at work or at home? Are you using chatgpt? Are you using Claude Sonnet? Are you using copilot? What’s your first go to Gemini, perhaps in Google. What do you go to usually?
Glenn Elliott
So I probably spend more time than most working with these tools. It really is just a nearly all day, every day phenomenon for me now, there’s a distinction between what I use to execute my own work and what we use with clients. And what we use with clients is much narrower, and it tends to center around So Mel, just right from the outset, we partnered with Microsoft, and have made great use of the Azure Stack and the Azure open AI relationship. We’ve put all sorts of other models and approaches to the test, and have found that stack to be every bit as good on nearly all measures, and it just comes with that wonderful reassurance that helps us sleep better at night when we work with clients. But in executing my own work day to day, I have more freedom, and these are more transient. Ai, experiences, one of the things that’s really transformed my world is voice enabled. AI, yeah, there’s something about verbalizing your thoughts that really crystallize it, particularly when you get too many ideas bouncing around your head, and you think, I know there’s a path through this, but I just can’t gather it all in my head. Being able to have that conversation with another human is great, but if it’s a hard concept, they need to be seriously expert. And I find that very powerful. The chat GPT voice mode is strong in this I have a shortcut on the lock screen on my phone. It’s just one tap to start a conversation with someone that has access to the vast, magisterial of human knowledge ever acquired up until about the end of 2024 so I find that very powerful, and it plays to my mobility. I’m always on the go, and I love to run and to take long walks, and I hate being tied to a desk. So if you see me walking around, say the Botanic Gardens in Sydney talking into my phone. There’s every chance there’s not a human being on the other end of that conversation.
Rob Pyne
I can relate. I have that similar one touch voice mode activated chat GPT. What have we become? Rob? And I will actually just pop the air pods in and be literally talking while I’m walking like you, and literally people thinking it’s talking to someone on the phone, but actually, I’m talking to ChatGPT and asking it questions and actually just downloading thoughts. Because often you know you’ll be thinking about something, and you’re kind of processing your own thinking and talking out loud and just processing your thoughts, and then it prompting you to think more deeply on the questions that you’re asking. Yeah, I found it to be game changer as well. It’s just like a digital sounding board. So it’s working exceptionally well for me too. So you’ve really emphasized enterprise grade security here and privacy as being a key feature of anything people are doing. It’s, it’s clearly the thing people are thinking about the most, because people are saying you can’t use chat GPT because it’s, you know, your data is not secure. You have to use copilot because that’s within the environment of Microsoft. It’s within the Azure environment to protect that data from leaking out into being used by the provider. Can you just speak to that? And how do you position these guard rails when speaking with highly regulated financial planning firms that are really conscious of their data not being leaked out and being
Glenn Elliott
used? Yeah, absolutely, it’s critical, and sometimes it’s a source of paralysis for firms to just be so worried about the risks involved in this and unsure of the letter of the law, which, of course, is always a little gray at the edges, to the extent that they feel paralyzed and they often do nothing. It’s our role as technologists to help them get past that and to see what are often very viable and straightforward paths through it with just the right vendor selection, the right platform selection, the right settings enabled in the various tools that they’re using. Some of the emerging AI platforms have provided this sort of secure enterprise grade robustness by default. Microsoft came out of the blocks very, very early in this generative AI journey, and said, that is our commitment. We’re not using your data for any other purpose, but to serve your needs. We’re not training models on it. It’s not going anywhere. Here’s our policy, whereas others, it’s more of a. A an opt in scenario. Oh, okay, you’ve got to go in this setting, in this menu, and make sure you switch to ON, to make sure that we don’t use your data for any other purposes. So even that distinction of knowing which platforms default to what is helpful, and yes, there are tools meant for more retail usage or ChatGPT or a Claude, or a Perplexity or something like that, are more B to C oriented, there’s no question. But you can get some very, very similar functionality with just a little bit of piecing together of more enterprise oriented tools, just a slightly more creative approach, and you can arrive at the same outcome.
Rob Pyne
Yeah. Okay, so we’ve touched on the fact you’re working in this business. Meldus that you’ve you’ve founded, so tell us a bit more about it. Tell us how you’re currently working with advice firms or businesses in the professional services space, using their data to surface more intelligent insights that they would otherwise not be able to gather. I’ve been following you on LinkedIn. I’ve been tracking all your posts, and I can see you’re right at the cutting edge of how people are using their data sets to be really powerful at creating insights on how to run their business. And it’s why we actually chose Practifi back in 2017 because we were using the platform everyone uses, it seems, and there was just no data. I could have client data, but I didn’t have any business related data. I couldn’t run my business with lack of insight into what was going on in our business. And so it was the key decision point for us about moving to Practifi to sort of run a business first that happened to be in the financial planning space, as opposed to running a financial planning software that really wasn’t a business, tool wasn’t really a business. CRM, so can you comment on that? What’s Meldus doing and helping businesses sort of go that next level.
Glenn Elliott
Absolutely, it’s been a really fun journey, because as much as I absolutely loved my Practifi CEO experience, understandably, as a firm like that grows, you get more and more managerial in your day to day life, and just a little further away from the problems on the ground and solving them with technology. That’s to be expected, but now with Mels, I’ve had an opportunity to really get back to the detail, and it’s it’s just been fascinating again, with with this array of tools available to us now. And so first and foremost, we started with experimentation. We wanted to see what was possible. This partnership with Microsoft emerged very quickly, and we started to create an array of just little agentic AI products. The first one, as I say, was, was work floor. The second one was, was much more agentic, just aimed at just non technical users looking to get really insightful, deeper insights from their information, without the complexity of additional training and complex user experiences, but then we started to apply it to different data sources and to structured data and unstructured data, and to give it more tools and to apply these layers of personalization, we’ve built agents that do PII redaction, for example, to allow advisory firms to experiment with AI without running the risk of exposing anything sensitive to a client, so we redact it on the way through, and then you can run more experimentation with it with more of an open mind. That’s been really fascinating. These quality control type of approaches and oversight type of agents really just super powerful, slightly simple. I must say, these are not really rocket science, but it’s almost like, use the AI to create the AI agent, and then you use another AI agent to check the work of the AI agent. It’s almost like the snake eating its tail. So that’s been a really fun journey, but equally, at the same time, just providing advisory services. You know, I’ve been very fortunate going on the FinTech and wealth tech journey that I’ve been on now for a long time, both in Australia and very heavily exposed and driving in North America as well over a long time. And so just being able to provide that level of broad expertise and insight and slightly visionary leadership, hopefully to advisory firms they don’t necessarily have the access to that that they might like. So we found that really exciting. It’s a nice opportunity to be able to just kind of get back to the ground level and provide that sort of advice.
Rob Pyne
Yeah, great. So if you were advising a growing financial planning firm today, what would be your advice regarding their CRM and tech strategy, especially if they want to future proof with AI ready infrastructure. I think one of the key points you’ve made is about this open architecture platform. You must go with open architecture, albeit with enterprise grade security. What else would you say to a growing financial planning firm today that wants to set themselves up for the future of AI and what it can
Glenn Elliott
do? Yeah, look very hard at the openness of platforms, at the commercial arrangements of platforms. There are still some very kind of locked down, long term, complex commercial relationships in the market that frankly need a new lens over them, because the costs associated with hosting data and processing data continue to come down. And so where you see these complex, modular. The platforms that have been around for a long time. Firstly, have they changed much in that time? If you’re looking at your contract renewal for the next three year term, how much has this stack moved on since we last signed on? That really needs a clear lens over now. And if it hasn’t changed much in the last two years, then it really should have, or at least you’d have a very, very clear path for how it’s evolving now, in this age of faster technological change than ever before. So be wary of those longer term contracts. Be wary of proprietary lock ins to platforms that the broader technology landscape doesn’t understand very well. We used to care a lot about selecting platforms that the developer community would understand well, so that you had freedom of choice for the vendors that you might choose to work with. That’s now extended to consider platforms that AI models know well, because it’s an extension of your technical team, you can even have a conversation, have a conversation with chat GPT, and say, right, if I told you we were moving to this platform over here. Tell me what you know about it in depth. Tell me what you don’t know about it in depth, and you’ll get a very meaningful answer. And compare and contrast that to these more open platforms, where the AI models have already consumed the API documentation and all the community help boards. That’s very, very powerful, but also going into this with a really, really nimble mindset, that it is vastly quicker than ever before to try an idea, to prototype an idea, to prototype another one, all within the guardrails of a secure and robust environment that’s not too hard to get started with, but don’t think, what can We deliver in six months? Think, what can we prototype in a sprint in two weeks? In four weeks, and you should find that there are ideas that will come to life or they won’t, pretty quickly, and that turnaround time now is faster than ever before.
Rob Pyne
Yeah, your insights and your experience in this space would be invaluable for firms like ours and other business owners to talk to or just to help shape their own direction. So how do people get in contact with you? Glenn, if they want to talk about what you know and what they should be doing to try and position themselves to leverage the available technology now and the AI automation that’s available, how should they get in contact with you?
Glenn Elliott
Yeah, I’d love to hear from any of your listeners who we might be able to help. LinkedIn is a great place to connect with me. As you mentioned, you’ll see me posting there regularly, usually in running gear or gym gear or something. Yeah, hit me up on LinkedIn. I’ll connect to everybody. I’d love to start a conversation with you. When those conversations are in person, there’s always a question of, is this coffee a clock, or is it Shiraz o’clock? We can address that when we connect, but that’s a nice problem to have.
Rob Pyne
Yes, I know you’re a lover of wine, and named all of your releases at Practifi after a wine vintage. So listeners take note if you’re catching up with Glenn after what time about lunchtime, if let’s go for five each, let’s go five o’clock. All right. Now, I’ve known you for a while, Glenn, I’ve followed everything you’re talking about, and it’s been great to reconnect with you today and listen to the way you’re seeing, the way financial planning firms are taking advantage of technology, and how it’s shaping the way businesses are going to be more productive and and just get greater insights into their business to make better decisions. So I really appreciate you joining me today. Gleny from The Trusted Adviser community, thanks for joining us on the
Glenn Elliott
podcast. Absolute pleasure. Thanks for having me. Rob
Rob Pyne
you. Thanks for tuning in to The Trusted Adviser. I hope today’s conversation brought you new insights, inspiration for growing your business. If you enjoyed this episode, please subscribe on your favorite podcast platform, leave a review and share it with others in the industry, and don’t forget to connect with us on LinkedIn for updates on future episodes until next time, keep building trust embracing innovation and driving success in your practice. You.