VIDEO TRANSCRIPT:

Welcome everyone to this virtual roundtable from Insightly Impact building an AI investment firm, AI-ready investment firm. I'm joined here by Richard Daugherty who leads our wealth and asset management and public statement and Chris McDonald, Capital Markets Specialist at Amazon Web Services. Welcome and thank you for making time, Chris and Richard. A public statement and AWS have recently sponsored a survey titled AI-powered investment firm that surveyed 500 senior executives globally across a range of sectors from asset managers to fintech and across regions globally with strong representation from APAC, Europe and North America. One thing that is very, very clear in the findings was that 73% of respondents agreed that AI revolution has begun in wealth and asset management. They're already using AI for client interactions, for portfolio optimization and employee productivity. And with all the excitement around Gen AI, Agentic AI. Over the next three years, the firms plan to aggressively adopt these technologies. Now, while the potential is very clear, most of these firms are still in early stages of realizing its full value. And that's where I want to open the conversation today with our experts here. Richard, so let's start with the current landscape. Where do you see the forms today in AI maturity?

So listen, firstly, thanks for hosting us today. Amazing to be here with you and Chris. But let's sort of take a maybe sort of take a step back and understand why we're here, why we actually are looking at AI as a solution. Over the last number of years, last seven, eight years, it's been. tectonic shifts across global markets. We've had and we do have we've had a couple of wars and one in the Middle East one just on my doorstep here in Europe. We've had the uprising of inflation and definitely in the UK the stickiness to growth. It's very difficult to get out of into that into that growth mindset and into that growth trajectory. we've also had you know the changes in policy from across you know the world when it comes to the political landscape and obviously the uprising in more recent times in tariffs and what does that what does it actually mean what does it mean for for businesses it means businesses are really looking to find new ways of working, right? They need to find a different playbook because the old playbook is not allowing them to take their quantum leap into the white space, right? And as we think about that, so we think about the white space and the pressures around getting there, we're thinking around fee compression, margin pressures, data fragmentation, regulatory complexity, legacy. legacy tech drag all these elements are really you know not allowing businesses to to operate where they want to operate which is around growth around building scalable revenue generating businesses and how do they unlock that you know you know there The new technology out there right now, which is AI, allows businesses to start really unlocking ROI and taking that quantum leap into that white space. If we think about where businesses are, I think before I jump into the maturity curve, it's very interesting to know that 75% of COOs... in our industry right now are looking for guidance on how to execute AI. Roll back 18, 24 months ago, those same group were looking at how to pilot AI, right? So we're moving into execution. I would say from a maturity perspective, there's about 40 to 50% just to what you're saying, Ash, that are deploying pilots. They're running POCs and what have you. have you then we can have about 30 to 35 percent that are really looking at functional deployments maybe into a specific area in the operating model specific area in the departments then we have you know those the advanced firms that are the likes of you know maybe the the bigger firms with bigger budgets that are starting to really scale and integrate AI into into their ecosystem and those firms are you know roughly 20 to 25 percent so Yeah, yes, there's experimentation happening, but definitely there's also acceleration happening when it comes to execution. And I think as we accelerate into execution, it's making certain that we can prove value early on to get the momentum to then scale later on.

Very interesting, Richard. Thank you for sharing that. Chris, how do you see this then transforming the operating model, especially for advisors and client engagement efficiency, all those topics?

Yeah, great. And thank you as well, right, for being here. Great conversation. Looking forward to today and, you know, some really insightful points there from Richard. I think I'd like to your question, right, we want to focus on. How it's going to reshape the operating model of kind of wealth management firms and investment firms like from a client investment perspective, I think a couple of angles that you need to look at it and from the various market segments that we have. So if we go from retail banking through the mass affluent market segments up to the high net worth kind of segments, we're going to see a difference. We're going to see a shift or we are already seeing a shift in that in engagement model. Right. And that is due to we're seeing more personalization. We've all been talking around hyper personalization. from customer 360, right? And a lot of that is from these joined up data platforms that are able to provide insights. We're surfacing insights, we're giving nudges, right? And this is really being accelerated. from the kind of advancements that we're having in generative AI. And then what we're seeing with that is, right, that's being delivered more and more by digital channels, right? It's kind of that move on to real true hybrid advice, right? So, and I'll come to that a bit as well, right? So providing advice to kind of the mass affluent and even kind of below those kind of segments, right? Where you're able to put kind of those insights, nudges, kind of workflows into those digital channels. is starting to grow. But what else is kind of happening in this space is we're seeing a massive increase in the efficiency of advisors, right? It's an area that's been long talked about of how advisors spend too much time shuffling paper, whether that be writing up meetings, doing onboarding kind of work. They need to spend more time on that differentiated list so they can go from servicing 150 customers to maybe it's 300 customers, right? And the way that is happening in the workflows I'm seeing, I'm sure Richard would see something similar. It's like, it's like they're kind of what you almost think of simple workflows but they've never been simple to do but we're now working right so it's called summarization so being able to take a call that a customer that you just had with a customer and being able to just instantly summarize that and create action points for the advisor for the customer take that and next level on and it's meeting prep right firms generally have either individuals or groups that are spending a morning getting ready for meeting prep now we're seeing firms and automating that and the result of that is it's freeing up advisors it's giving them more time so they're able to spend more time more value add with the customers and potentially spend more time with newer customers that they can bring into that kind of set that they have with them and i just quickly say i think on the operations piece right so outside of the advisor piece we're seeing big steps right between onboarding surveillance onboarding has you know long been taken kind of advantages of ocr for kind of doing the onboarding extracting information and data but we've seen great improvements in that with the capabilities of large language models and kind of that summarization and extracting kind of sentiment so an improvement on that side and then from the surveillance side right so from both a post-call analytics and even in-call kind of live agent assist so when an agent or advisor is on a call giving them details like what they should be talking to the customers about but also from a From a compliance perspective, are they providing the right disclaimers around certain products? And then you take that to the next stage, analyze all those calls, right? So you can then... ensure that those conversations have happened. So I think we're going to see a big change, or we are starting to see a big change in a couple of areas, like more engagement through the digital channels and the advisor. It's definitely not disappearing. The role is improving and they're working alongside AI. But that kind of autonomy piece that we talk about, there's autonomy in the background of seamlessly joining up apps through kind of API connectivity and MCPs, but the human is... And it's very much in the loop still.

So more enabling the human and bringing value to the customer. That's fantastic.

Yeah, 100%.

Yeah. What are the key challenges which are you seeing firms face while trying to scale from those experiments that you mentioned earlier on towards full scale implementations?

Yeah, good question, Ash. So a couple of things. I mean, I think a few things. I think one key ingredient in recipe is making certain before you start that there is a very, very clear understanding. Right. So we're not talking about ROI. I'm talking about, you know. Are we able to, in this pilot, mitigate risk? Are we able to potentially become more cost efficient? Are we able to drive revenue? We need to look at those elements and really understand how we can prove them. in a pilot phase in a POC phase of near six to eight weeks and you know are we going to get all the way there no we're not going to get all the way there 100% not but we're going to be we're going to be able to create the foundations that allow us to build the momentum and also the confidence that if we then go on to a 12 or 18 month journey that those foundations are embedded that value will then be released into the ecosystem depending on where you choose the value to set, right? So that's one part and it's critical that there's always an iron ROI and there's an iterative view of how to achieve ROI. I think that the other thing is, you know, there's also some groundwork, one-on-one things to remember. Having execution discipline is still required, by the way. You know, just because we run in a new playbook doesn't mean we should change some of our execution habits. We need to understand the business. We need to understand the requirements. We need to take the business users, the technology users, the whole organization along the journey. We need to communicate. All these things need to be in place in order for us to achieve what we want to achieve. Execution discipline is, you know, top of the list, right? Then there's, you know, some technical elements that obviously, again, are required. of foundational elements that need to be achieved in order for us to actually release the value of AI, we need to make certain that, you know, our data and legacy systems aren't fragmented. You know, still today, there's a lot of unstructured PDFs, emails, legacy platforms that's preventing single source of truth. And, you know, if we look across some of the markets. So AI is helping drive data reconciliation and cutting sort of manual effort across these elements. So we need to, as we start off, make certain our foundations are set up and we're not building on sand, for want of a better word. Then I guess just to wrap it up, I think the important part of scaling is making certain that. We view AI not as a technology problem and just because we're hiring different people into the organization that have AI ahead of their title doesn't mean it's only their job to look at AI. It's everyone's role, right? It's everyone's role to take a look at the problem, take a look at the new solution out there and work together on how to scale it. And I think... That if we're able to orchestrate this along with the agents, we're able to then move into a space where we have data, we have embedded transparency, and then we have ROI as the anchor that allows us to release the value into the organization, into the ecosystem on an iterative basis.

Fantastic. Yeah, I mean, investing into the foundations sounds like a really good thing to start with. Chris, what advice would you give to firms that are stuck in pilot? I mean, I hear about a lot of firms that initiate a pilot, but then some are unable to come out of the pilot mode and not scale it. So how do they create value for the whole enterprise?

Yeah, you know, I'd probably echo some of Richard's points in his last answer as well. But I think first of all, right, it's like getting out this POC mode. We're seeing more and more firms get out that POC mode, right? When this kind of started three years ago, everyone kind of like starting their chat GPT on the phone, then you've got 100 use cases, and then we're getting a bit smarter, and we're building kind of some rag infrastructures, but we're still because you can is that technology, it's not like, where you need everyone involved from a blockchain perspective, kind of that. of that kind of build with Gen AI everyone was building POCs and maybe not the smartest but I think we're getting a lot smarter about being those POCs so it was allowing you to build a POC but fail quickly and then start again right a really iterative agile kind of product building kind of mindset but what I what I would say to ensure success and that we're able to do this first of all right I think everyone knows this but you need a modern data infrastructure right so That means investing in the cloud. It means building data lakes. But more than that, it means it's a data warehouse and it's potentially a data mesh, right? It's that kind of integration, being able to build that data platform that allows for consumers and producers to access the same data and you have data lineage, you have data audit trails, you're all pulling from the same bucket. And why that is so important is because everyone is working with vast amounts of data and that data is only increasing, right? Whether it's pulling more information from around your firm or you're accessing. interesting new data sets whether that be else and then on top of that right in order to be that differentiator to do something different to the other firm firms are building these knowledge bases and rag infrastructure so retrieval automates and generation right to be able to do that so super key to get that data platform right and as important which sounds obvious is build the Gen AI machine learning platform, be strategic about it, don't be tactical and have little builds here and there. Be clear from the start on what you're building and make sure everyone in the firm is built in. And that is coupled with at that same time, focus on governance, focus on controls, build those guardrails, have the security controls. Like Richard said, you're still building product, right? It's got to have the same checks, balances in place as you have with building any product over time. So embed. governance early and bring everyone along for that ride right so as you're building so and with that start with strategic workflows and what i mean by that is like to what richard was touching they need to be measurable they need to be quantifiable results so you can demonstrate their roi so you know and where i've seen firms have success is like i said once i said earlier maybe you start what you think is smaller but like in the summarization and the meeting preps in the one-to-one personal kind of marketing that you can do by creating insights and nudges and why they're really good you know it's you can identify the impact right suddenly it's like it's not taking me three hours to summarize a meeting or my advisors got so you have to do that and by doing that and if you've started the governance and all the kind of steps that need to take place from an early stage you start building that culture in the firm as well because you're bringing along compliance you're bringing along the advisors you're bringing along products you're bringing in engineering so everyone's working together and I will add to that right so being strategic doesn't mean starting with the simple workflow right because some people have seen do that say start simple and it's something you don't need you can think complex but set that North Star and iterate through it right build what makes sense you know you're not going to suddenly change the way you build product so I think you know if I was to summarize that it's get the data and your machine learning Gen AI platform correct from the start ensure governance is in place and build strategically

Fantastic. And I think both of you actually talked about ROI and value coming out of these investments. Richard, what does ROI from AI actually look like? And how do you recommend that firms track it and measure it?

Yeah, I mean, listen, I think it's critical and, you know, as Chris was alluding to, it's one of the cornerstones to any project as we move forward. So I think, you know, good ROI from an AI point of view isn't just about, you know, embedding orchestration agents for the sake of embedding orchestration agents. We want to make certain we are. or achieving measurable value across cost, risk and also revenue. If we think around cost, cost efficiency, how can AI deliver 20 to 50% cost efficiency across, you know, manual workflows? Decommissioning EUCs is a good one. There's a nice use case that we're working on on EUCs. Also, maybe potentially also shortening IT chain cycles. Of course, you know, I'll give you an example here. We're working with a team on a guideline monitoring agent where we're able to insert an agent into an operating model that has now meant that 200 hours of work has been cut out of that team that allows now, it gives the team, the MD and the team the option.

within his team and therefore they can grow that business much much quicker if they want to but now here's the option six months ago he didn't have that option right so um that's that's interesting and then obviously if you look at you know risk mitigation making certain that you know chris alluded to it from a compliance and risk reputation risk perspective there's elements already up and running that um that uh like agents can can really take uh take a note of right compliance screening arises gaining is another one I remember running a few projects a number of years ago and so many different so many different elements were changing on a on a monthly, quarterly basis. And it was difficult for the program team to keep up because, you know, there's fluctuation in the requirements from the regulators. And obviously then the compliance team were not able to keep up. So that meant there was sort of wasted effort and wasted time and wasted money spent. But if we have right now, we have an agent, we have a rising scanning agent that can really simulate and understand. What's going on from a FCA, SEC, BOE perspective, it's able to then relay that back into the compliance team and gave really good up-to-date information back into the business and act as a strategic pointer. So clearly reducing the risk levels inside the business. Revenue generation, you know, we obviously see Chris. touch on this upfronts we're wanting AI to and agents to work alongside as co-pilot inside investment firms and they are I mean we've definitely we've seen this already we're building out investment research agents that are releasing early value into into PMs those PMs then then can they can then serve as multiple clients because they've got the they're getting information in much higher speed and those clients can obviously then you know access different different information different funds different different positions at a quicker speed so you know it's it's critical to to look at it from that lens as well but I think you know just to wrap it up there's making certain that We look across those three lanes, but also as we decide on which pilot to go with or which use case to go with, there's the business value assessments. So where we sit in from a cost, revenue and risk perspective, but also feasibility point of view. So is this complex, highly complex? Is it not complex at all? And what is the delivery risk associated to that? that. So again, going back to wanting to deliver early value into an ecosystem, you want to assess all these elements so that when you're delivering maybe your first use case into your ecosystem, some of these parts are mitigated so you can release that value. And then you can get maybe onto the complex items as the board, the executives gather confidence and the firm gathers momentum. in some yeah

Fantastic. And big question, guys, Chris, to close, how would you characterize leaders in the next few years? So what would a leading AI firm should look like?

sure um probably gonna touch on again some of the points we touched on but i think

First of all, the successful firms are not going to be defined by firms who've just built AI. It's going to be those that have built around it, meaning you've built intelligence infrastructure, right? So the whole kind of data conversation we've touched on, right? And being able to take the machine learning, generative AI platform and being able to iterate and take all that. I mean, obviously the space is moving so fast. So being able to build and add in the new kind of... technological capabilities as they kind of progress so having that adaptive platform that is able to kind of keep up with the momentum that we're seeing and make no mistake data is going to be that differentiator right is kind of how you really build out the super powerful platform and these platforms are going to be we're going to have more assistants, more agents working alongside advisors, right? I think that's really key. We're going to see areas where we're going to have direct-to-consumer capabilities, but where the advisor is being able to work alongside that generative AI, right? So whether it's meeting summarization, there'll be some parts that are fully automated that they'll be able to pull through, but really working together. And there's going to be an increase as well, right, in these platforms, in these kind of multi-modal. the modal platforms that we'll see because this space is converging right you're seeing banking step into wealth management it's just having a savings account it's no longer enough private banking is moving into this space we already have obviously the wealth management areas but retail brokerage trading is coming into this space insurance is coming into this space right everyone is kind of converging together to build this holistic wealth management kind of service and single pane of glass i think we're going to continue to see down that part We may see, and we're starting to see in areas like more conversational chatbots, even avatars kind of come into that space, right? So we're kind of starting to see that, you know, it's kind of a bit more fun on that part. But I think, you know, key as well not to get carried away. There's still going to be human in the loop. The advisor is needed there. I mean, who knows two, three, four years out how this continues to progress. But really. Those advisors are going to be able to, right, not just the advisors and actually kind of directly, we're going to see more people receiving wealth advice and receiving better advice, right, from these tools. So it's a growing space and from all the kind of indication it's needed right from the intergenerational wealth transfer that's happening, more wealth out there to be managed, etc. Right. So that's kind of going through. And I think we'll see kind of this real development in hybrid advice, right? So robo advice, no, but hybrid advice. elements we're going to get it from the direct from the platform but when you can you can talk to an advisor right so that advisor also taking more of a proactive role as well so this kind of idea of continuous advice rather than just a semi-annual annual kind of check-in being able to put that kind of information out and kind of working with them but key to all of this success is the platforms have got to have the governance and they've got to be explainable right so in order to build that trust this is a trust space um that we're working on and i think i just close one other thing and i close and i think we're going to see we will see firms continue to move down this path and because customers are going to expect it right we're already doing these kind of workflows and other businesses and other areas you know what we're kind of doing so there are going to be those expectations so i think You know, a couple of points. The advisor is not going to necessarily be replaced by Gen AI. Maybe the advisor that doesn't use Gen AI will be replaced by the advisor that does use Gen AI. And the firms that really succeed, it's not just going to be better technology. They'll have reimagined how technology works with human expertise and how they work together.

And just to, just, I think those are, you know, incredible points just to add to that. I really passionately believe that we are at an inflection point here where AI stops being, you know, stops being a project and becomes really the pulse of a firm and the leaders who are taking charge now and scaling this will, they have the opportunity to define. industry in 2026 and beyond and I think that is really really exciting and just just adding to what Chris was saying which I think is absolutely amazing

Definitely a very exciting time to be in. And thank you so much guys for your time today. It's been an amazing conversation. We'll be sending out a copy of the report so everybody can actually get more of these insights and some case studies on what leading investment firms are doing. and obviously to our listeners right so if you if you or your team members want to discuss any further both we at PublishSafe and as well as Amazon AWS are very happy to engage with you we look forward to the collaboration so give us a shout for any of these topics we've been working together for a long time PublishSafe and AWS and we've been a long-term partner delving into end business transformations and we experience delivering end-to-end on emerging firms as well as using the best of cloud and AI solutions coming from AWS side that basically brings the magic together. So, like Richard said, we are at an inflection point. AI is not just a technology, it's a catalyst for reinvention and we look forward to working with you. So, yeah, let us know of your thoughts and share your comments with us. Thank you so much, everyone, and thanks especially to Chris and Richard for your time today.

Relevant Links

• Building the AI-Ready Investment Firm: From Pilots to Scaled Business Value
• Wealth and Asset Management AI Execution
• Measuring AI ROI in Wealth and Asset Management
• The future of wealth advice will not be defined by robo-advisors replacing humans.
• AI in Investment Firms: Trust, Transparency and Resilience
• De l’expérimentation à l’exécution : comment les sociétés européennes de gestion de patrimoine et d’actifs peuvent réellement industrialiser l’IA (Europe)
• De pilotos a escala: cómo las firmas de inversión en México pueden convertir la IA en una ventaja operativa real (LATAM)
• FAQ (FAQ)
• What to Know About Building an AI-Ready Investment Firm: 10 Key Takeaways for Wealth and Asset Managers (LIST)