Watch on YouTube     |     Listen on Spotify     |     Listen on Apple Podcasts

"The proliferation of AI agents performing a number of tasks for us in and around moving value means they will prefer to use, and execute, via wallets, blockchains, and tokens. That trend broadly put together represents better, faster, and cheaper transactions. Cheaper in terms of transaction costs, faster because everything is in real time with atomic settlement, and better because they are inherently more programmable, both at the smart contract layer and in their interaction with AI agents." - Richard Schroder

"The first place to look is the existing real-world assets, which is around 20 trillion dollars, while what is tokenized to date is only around 38 billion. By any measure, we are at the beginning of the journey. The adoption speed depends on how fast we can show proof points around why being in tokenized form creates additional utility, alongside the speed of what is happening in AI. I believe that agents will prefer to interact with stablecoins, tokenized real-world assets, and crypto rather than interacting with the traditional fiat banking system." - Richard Schroder

"We are bringing in Australian and Singaporean money market funds because, in this part of the world in APAC, we are a multi-currency jurisdiction. For the most part, there are only US dollar stablecoins and money market funds available, and we do not think that should hold true for APAC over the longer term. We think it is today a multi-currency jurisdiction in trade and capital flows and will continue to be so, which is why we are building for that with our multi-currency capability." - Richard Schroder

Money is being rebuilt for AI agents, and the first regulated funds are already open to them. Richard Schroder spent 25 years in institutional banking, led the ANZ digital assets team that issued the first bank stablecoin on a public blockchain, and left to co-found Conduit Digital Holdings, which tokenizes money market funds for corporate treasuries across Asia-Pacific. He joins Jeremy Au to cover why asset managers moved on-chain before banks, what Sibos and Token2049 now agree on, how a tokenized fund is made agentic-aware, the three controls that let an AI agent subscribe, redeem and transfer without going rogue, collateral mobility, how Federated Hermes and BNY fit into the stack, where the moat sits when most of the technology is open source, and the gap between a roughly $20 trillion real-world asset market and the $38 billion tokenized so far. Plus his advice for fund managers ready to make a start, and the $30 million stablecoin trade that felt brave at the time.

Richard Schroder is the Co-Founder and CEO of Conduit Digital Holdings, the digital assets business of Conduit Group, an independent investment group headquartered in Singapore. Conduit Digital Holdings tokenizes regulated money market funds for institutional and wholesale investors in Asia-Pacific, with a strategic alliance with Federated Hermes, and offers tokenization-as-a-service and agentic services to fund managers. Richard previously spent 25 years in institutional banking and led the digital assets team at ANZ.

Richard’s LinkedIn: https://www.linkedin.com/in/richard-schroder-76961140/ 

Conduit Digital Holdings: https://www.conduitdigital.xyz/

Want to get to the latest updates, access to transcripts and exclusive event invites? Sign up now - It's free!

Money Is Being Rebuilt for AI Agents. Where Do You Start? | Richard Schroder - E745

From Institutional Banker to Founder

Jeremy Au: Welcome to BRAVE Southeast Asia. I'm so excited to be here with Richard Schroder. He's an expert, a leader and an investor in digital assets, and in understanding what needs to happen for everything you want to know about real-world asset tokenization. This is a wild frontier for me, and I think it'll be a frontier for many of you as well.

I'm really excited to talk about this because we're seeing a huge transition from regulated institutions, the traditional approach of finance with its suits and ties, to this wild world of decentralized finance. I'm so excited to have you as a guest here.

Richard Schroder: Thanks, Jeremy. It's great to be here. I'm a family guy. I've got four great kids, a neurotic dog and a beautiful wife. We live in the suburbs of Melbourne, and that takes up a fair bit of my time. In terms of my career, I was an institutional banker for 25 years, in project and structured finance through to corporate finance and corporate treasury solutions.

But around 2014, I got really interested in the exponential technologies, the technologies that double in performance and halve in cost every year, and tend to reshape industries. So I started gravitating towards them and took up roles leading data analytics for the institutional division at ANZ, and then leading the digital assets team at ANZ. Then I spun out with my own new venture, Conduit Digital Holdings, in partnership with Conduit Group and, on the tech side, Cubewire. We're about ten months in.

Why Asset Managers Moved First

Jeremy Au: Wow, fantastic. Lots of growth there. What's interesting is the timeline of that transition from traditional finance to, let's say, the brave new world of digital and decentralized finance. Could you talk a bit more about what you saw that made you say, "I could keep going this way, or I could do something different with the technology and the approach"?

Richard Schroder: The big transition for me: we pioneered in this space at ANZ with a stablecoin in 2022, the first bank to issue a stablecoin on a public permissionless blockchain. They say being early is sometimes the same as being wrong. There just wasn't the customer demand at the time, or the risk appetite. Banks in particular have to work within prudential regulatory frameworks, and those just haven't been developed enough for them to operate on public blockchains.

It was the asset managers who were the first regulated players to move onto public blockchains, notably BlackRock, Fidelity, Franklin Templeton and others. That was around 2024, and it inspired me to get out of banking and into asset management.

Sibos and Token2049

Jeremy Au: So there you are, having seen both sides of the table, traditional finance and decentralized finance. What would you say are some of the differences, before versus after?

Richard Schroder: I've just been on the conference circuit. The largest global banking conference, which happens every year, is called Sibos. It was in Miami, with 12,500 bankers, central bankers and regulators gathered together. One of the main themes of that conference was the intersection of tokenized finance and agentic AI, and the role it has to play.

Then I come down to Token2049 here in Singapore, and there are 25,000 crypto bros in T-shirts. And what do you know? The main theme of the conference this week is the intersection of tokenized and agentic finance. So there is a convergence happening. But I don't think there are many companies at both conferences, so there's still a bit of a bridge that needs to be crossed. And at Conduit, we seek to be that bridge.

Jeremy Au: What's the difference in flavor in how these two conferences approach the same topic?

Richard Schroder: In banking, it's always about the regulatory parameters within which the banks can operate. There's general agreement that the question is not if, but when, and how to build the capability so they have the strategic optionality to move when they need to, and when their customers move. So there's a lot happening where they can be on blockchain: private permissioned blockchains, tokenized deposits and so on.

Token2049 was very bullish. A big theme is financial instruments, RWA, coming on chain. That's part of why there's that bullish sentiment, because it's using the crypto rails. By volume, the traditional market is still significantly bigger than crypto, so as those assets start to come on chain, it's going to be big for the underlying crypto rails.

Better, Faster, Cheaper, Programmable

Jeremy Au: We're going to talk about the trend, and eventually we'll talk about Conduit. The trend we're talking about is the tokenization of real-world assets as digital assets. For someone who isn't familiar with this trend, how would you describe it?

Richard Schroder: There's an inevitability to it. There are the rails themselves, the use of blockchains, and then the smart contracts that can be executed on chain, the tokens that can be held on chain, and, increasingly, the AI agents that can interact with those tokens and wallets.

Put together, that trend represents better, faster, cheaper. Cheaper in terms of transaction costs. Faster, because everything's in real time, with atomic settlement between assets and the like. And better: the big one I see coming is programmability. As AI agents proliferate, performing a number of tasks for us in and around moving value, they will prefer to use and execute via wallets, blockchains and tokens.

From Stablecoins to Money Market Funds

Jeremy Au: It's interesting to see the different types of assets being moved and tokenized. We've obviously seen US Treasuries. I guess stablecoins are a form of currency, but also of tokenization. We also see student debt and other kinds of assets. Can you talk more about what kinds of assets are being programmed this way?

Richard Schroder: Stablecoins were the first, as a medium of exchange. Then the money market funds. I didn't see that one coming. Money market funds are amongst the most efficient financial instruments going. There's nothing wrong with them, so why do they need to be put in tokenized form? Initially, it was to provide access to a customer segment that preferred to use tokens, namely the digital asset natives: the exchanges and the projects.

But since then, additional utility has really started to come through, and that's supporting further growth into other asset classes. That utility falls into a couple of areas. The first is collateral mobility: the ability to get more out of your asset, to use it, lend against it, move it around, and use it as collateral for trading and the like. The second area is programmability. Tokenized assets are inherently more programmable, both at the smart contract layer and in their interaction with AI agents.

Jeremy Au: Why is programmability important from your perspective?

Richard Schroder: It's at the heart of productivity, of automation, of making better decisions, supported increasingly by AI, and of enacting those decisions through agents. That's the key area. What about you? Are you playing around with AI agents? Have you got any portfolios going? What's your take?

How Jeremy Uses Agents

Jeremy Au: For me, it's primarily in two domains. The first is personal productivity and learning, and the second, of course, is work. On the personal side, I read a lot. I have my little agent gathering a bunch of desktop research on the trends I like. I used to use an RSS feed, and it would give me 100 links to click through and read. It's been nice to have something run around, do a quick summary and let me dig deeper. It's also contextualized by me, for example when I ask questions at night about what's happening in the news or in geopolitics. So there's a nice personal learning side.

On the work side, it's just some very basic agents around emails and very basic research for meetings. I haven't done the work to let it actually run a business function yet. Personally, I'm waiting for the right platform that makes sense and won't do anything irreversible. You've seen some of the scary mistakes people have made. Recently, I think somebody's barbecue meat shop just emailed its entire customer base. So it's interesting to see.

Opening a Regulated Fund to AI Agents

Richard Schroder: Yeah, it's a really interesting area. I'm a director of the fund, so if some agent goes rogue on my fund platform, I could go to jail. That's the last thing I want. We've actually made our funds available to agents already. That's called being agentic-aware, and I think we'd be amongst the first regulated funds to put that in place.

We worked through a few key due diligence areas, because we have a fiduciary obligation to our customers. We have to assess a platform before we allow it to integrate with our tokenized fund platform. Out of interest, and for your audience: beyond the information security review and the normal stuff like that, the three key areas we came up with were these.

First, a unique agent identifier, and how that's linked through to a KYC-onboarded customer.

The second was the ability of the platform to define the agent's tasks. There are only a few tasks you would expect an agent to be doing in a tokenized fund environment: subscribing, redeeming, transferring. They shouldn't be going anywhere near the whitelist contract, or other things like that.

Finally, and probably most importantly, the controls that can be put around what the agent can do at the wallet level: threshold limits, and multi-approver capability where you want it. That puts the user back in control of their agent.

Collateral Mobility

Jeremy Au: So what are the benefits for your fund of being tokenized, and also of having this access for agents?

Richard Schroder: It's a great question. We're starting with multi-currency money market funds. As I mentioned, the underlying funds we're tokenizing are amongst the best in the world. What we're seeking to do in this first phase is demonstrate to ourselves, our investors and our customers the additional benefit of being in tokenized form. For us, in the initial phase, that's about collateral mobility, getting more out of your assets, for example the ability to lend against them. And secondly, programmability.

A good example of collateral mobility that's come through really recently is the tokenization of US stocks. That puts owners of those stocks in a position to readily access liquidity or lend against them while holding that long-term position, maybe to get a bit of working capital for something they want to do. So it's getting more out of their asset, and being in tokenized form really enables that.

Corporate Treasury as the First Customer

Jeremy Au: What's the breakdown of the people who are excited about this? Is it, for example, crypto bros, as you mentioned? Is it family offices? Who's making the decision to say, "Okay, this is something that's really compelling to me"?

Richard Schroder: For us, our target customer segment is corporate treasury. Within that, the starting point will be the digital asset exchanges, through to mid-market corporate treasury management. It's about them being able to really reshape, and make much more efficient, their ability to go in and out of positions.

A typical flow for an Australian company, as an example, is that they'll earn in Aussie dollars but buy in US dollars. They want to stay in safe, secure yield on both the Aussie leg and the US leg, being able to move between the Aussie money market fund and the US money market fund. Then they can use the best of agentic capabilities to forecast what's coming up, prepare for it, make those movements, pay suppliers, and just operate a lot more efficiently and effectively.

Jeremy Au: The value of tokenizing the fund makes total sense, because it gives people more optionality to move fast. I'm always curious about who's at the other end of that. Is it a CFO making the decision? Is it a platform making the decision? Is it an engineer? And then they send those agents to work with your fund through the various middlemen and platforms?

Richard Schroder: I think it very much has to come from the C-suite level. They're effectively onboarding a form of extra staff, so it's about putting the right governance around it. First the business requirement, then the governance around it, and then it passes through to the tech teams to execute. But we are talking about a future state here. As I say, I can't find other regulated funds that are available to agents yet.

I think where we are with agentic is more like what you describe for your own personal use cases. It starts off with generating some interesting pictures, then it goes to supporting your information needs, and then ultimately to deploying agents to support tasks. In a corporate environment, I think we're still talking about a future state. There are a lot of internal use cases deploying agents to support engineering functions, but what I'm talking about is corporate treasury. That has to come from the business, because there's a lot of governance overlay and there's value being moved.

How Traditional Banks Are Responding

Jeremy Au: How are traditional banks reacting to this?

Richard Schroder: As I mentioned, coming from Sibos, there's agreement that it's a question of not if, but when. They're mobilizing where they can, particularly in private permissioned environments and with tokenized deposits.

We have a great bank that supports us in BNY. They're the largest custody bank in the world, and they're one of those forward-looking banks that is, for want of a better word, getting around this. As the largest custody bank in the world, they're already managing the largest amount of real-world assets, and they're making those available for tokenization in an appropriately regulatory-compliant but also efficient way. So firms like ours can access the underlying assets, tokenize them and get them distributed to our customers, who can then take advantage of some of the things we've talked about from being in tokenized form.

There'll be other banks who will try to resist the change as much as they can. The better ones will seek to reposition their business in anticipation of it, and take advantage of the technologies themselves.

Jeremy Au: How do you pitch to those forward-looking banks?

Richard Schroder: Well, I'm an asset manager now, so the bank's serving me, and we've got a great one in BNY. But when I'm talking more broadly to banks, fund managers or existing players, the starting point is that you've got to get a wallet, a wallet that enables governance around it. And you've got to start finding use cases within your business that might take advantage of some of this new technology, and then iterate from there. How do you eat an elephant? One bite at a time.

Why Partnering Beats Building

Jeremy Au: Is it easier for the banks to build, buy or partner in this brave new world?

Richard Schroder: I think partnering is critical, and it's certainly at the heart of what we're doing at Conduit Digital Holdings. TradFi does certain things really, really well. The first fund we've tokenized a portion of is the Federated Hermes US Prime Fund. We're very proud to have Federated Hermes as a partner on this. Their underlying fund has hundreds of credit instruments that they manage really well. We don't need to replicate that, and we don't want to replicate it. This is a safe, secure, well-managed triple-A fund. I've talked about BNY as well.

Then there are certain things DeFi does really, really well. Some of the protocols that came out of DeFi Summer have been in operation for over six years, and they're very much battle-tested, such as a lending protocol or a market venue. I think the secret sauce is in configuring the best of both of those worlds.

What Makes a Moat

Richard Schroder: We're raising some venture capital at the moment, and "What's the business moat?" always comes up. There's a tendency sometimes to look for the patent, the particular one-off thing that separates you. My view is that it's actually the configuration, and the speed and quality of delivery of that configuration. Across the business side, the partnerships and alliances you form. The legal and regulatory side, the licenses and jurisdictions. And the technical side, where most of it's open source, so it's how you bring it together in an institutional-grade stack, something that's well certified: SOC 2 Type 2, ISO 27001. Put all that together, deliver and execute well, and to me that starts to look like a business moat.

But you're the venture guy, so I want to ask you the same question. What defines a good moat, and how would you look at this space?

Jeremy Au: Historically, moats ranged from weak to strong, and from well understood to less well understood. IP and patents would be one. Number two would be economies of scale. Another would be a team in a limited talent world. For example, quantum computing has a limited pool of talent, so talent is a moat. Regulatory licensing is a big one. And the last would really be network effects. I'd say network effects are probably the strongest moat of all in the digital world, but also probably the least well understood.

Richard Schroder: And the hardest one to get.

Jeremy Au: Yeah, it's hard to get.

Richard Schroder: Particularly at an early stage.

Jeremy Au: Yeah. Some things are easier and some things are harder. But to get a world where every additional user makes the product better for everyone else, and to have that at the company level, is difficult. I think Tesla is a good example. Every person who buys a Tesla trains its computer vision for its self-driving models, so the self-driving cars get better and better. Whereas Waymo at Google has to buy every car itself to get the training data. So it's interesting to see those moats. How about you? When you think about it, I think regulatory is of course a big one. Compliance, leadership.

Richard Schroder: In my regulated financial instruments context, I see it as that configuration across the business, regulatory, legal and institutional-grade tech. That's what matters when you're moving value around, which is a bit different from when you're just moving information around.

Moving Information Versus Moving Value

Jeremy Au: Let's talk about that. What's the difference between moving information and moving value?

Richard Schroder: Moving value has traditionally been very much the domain of banks. Now, through blockchains, there's a considerable amount of value being moved independently of banks. But it's really in the quality of the controls: the cybersecurity controls put in place to support the stack, and how the engineering code gets released and deployed.

It's a shortcut way of thinking about it, but there's such an important role for certification and standards. That's where SOC 2 Type 2 comes in: how you actually run your processes end to end, and getting that attested. Then the ISO standards, which reflect how you go about your information security practices, the likes of ISO 27001. Put together, those reflect discipline around deploying your applications.

When it's just a website with information flowing, there can be some important customer confidentiality considerations, and that pushes the need for something more institutional grade. But if it's just an information exchange website without any particular consequences or private information, that's a lesser standard to adhere to. Most of those information-based websites then have you click through to a financial service, be it the likes of Stripe.

Stripe and Agentic Commerce

Jeremy Au: That's interesting, because historically people think of Stripe as a platform that helps people with their e-commerce transactions. Now you see them starting to expand into providing treasury services for companies as well, doing all these things that look a little bank-like. I'm curious how you see that trend.

Richard Schroder: I think they're going along exceptionally well. Alongside others, I think they're leading the way in this first phase of agentic commerce. But I've not seen it elsewhere yet, say in the regulated fund space, which is the path we've gone down. So I still think it's early days for what they're doing in that space, but they're showing the way and it's very promising. What's your view on where things are at in the evolution towards agentic commerce, or the use of blockchains, by those players?

Websites Built for Agents

Jeremy Au: Well, I'm somebody who likes reading science fiction, so where my brain normally goes is this: I'm pretty sure an agent, maybe even today, is out there buying server space for itself to continue living an independent life, doing the transactions it needs using blockchain to sustain itself, generate economic value and stay alive. That would be my science fiction side.

That comes back to what you just said. I think there are going to be a lot of people deploying agents at the enterprise level, or even as sole proprietors. Even today, I see a lot of engineers busy building their own agents to look at their personal finances. For myself, it's primarily been on the information side, using agents to understand information for finance decisions. But I'm not letting it touch my wallets or my bank accounts, anything.

Richard Schroder: It's coming.

Jeremy Au: It's coming. I think there'll be a day when that happens, and I'll probably pay somebody to take on the liability of what happens if it goes wrong. But it's coming, so it'll be interesting to see how that plays out. I was just reading today that Cloudflare reported that, of human traffic versus bot traffic, the internet is now majority AI traffic. So it's quite interesting to see that happening.

Richard Schroder: Yeah. I love showing an MCP server, the Model Context Protocol. It's ultimately reflected as a website, but it's not like a human website. It's not a nice UI that a human engages with, with buttons and dashboards and the like. It's just a whole lot of text that reflects the services available on the website. It's the same sort of thing, but a very different user interface, because these large language models are text based. A UI that's good for them is not the same as a UI that's good for humans. So what you're going to see, with this traffic, is a majority of websites that look not so great to humans but beautiful to agents.

Jeremy Au: Yeah, it's a wild world. One point Cloudflare also made in its comments is that the majority of traffic on the web today is now AI bots. That's great for the bots, because they're out learning and reading everybody's Tumblr and WordPress accounts. But the humans who maintain those websites and pay the server costs for that traffic always assumed a model where humans are reading. A human will read maybe one or two pages of your website. A bot is going to read every page, every graphic and every hidden subdomain you forgot to lock down.

So there are actually some cost dynamics for webmasters and engineering teams, who have to ask whether they're making their websites for humans only. Some have tags saying no to bots. Who knows how effective those warnings are? Others are open to both humans and robots, some are charging robots, and like you said, some will probably become robot-only. It's going to be wild.

Richard Schroder: Yeah. Going back to regulated financial institutions making their platforms available to these agentic platforms, there are some questions to be asked of those platforms about whether they're suitable for integration with a regulated financial platform that's going to result in the movement of value.

Jeremy Au: That's interesting, because we already see agents online repurposing message boards into forums to trade information. But you're right that value is something that's difficult for them to access as of today.

Conduit's Products

Jeremy Au: So let's talk about the products Conduit has.

Richard Schroder: We're multi-currency money market funds to start with. The anchor product is a tokenized version of the Federated US Prime fund, and then we're bringing in Aussie and Singapore dollar money market funds. In this part of the world, APAC, we're a multi-currency jurisdiction, yet for the most part there are only US dollar stablecoins and money market funds available. We don't think that should hold true for APAC over the longer term. APAC is a multi-currency jurisdiction today in trade and capital flows, and it will continue to be, so we're building for that with our multi-currency capability.

Then we'll go up the risk curve. Stocks are a really interesting one. They're the flavor of the month, following the SEC's innovation exemption allowing for the tokenization of US stocks. There are significant flows of retiree money in particular out of APAC into US stocks, and my home market of Australia is a big one. So we'll have a look at that. Commodities and private credit have already got off to a good start. But we'll be led by our customers in terms of what comes next. We've got to get across the main currencies for our core multi-currency tokenized money market fund offering, and then we'll move up the risk curve into other tokenized RWA.

A $20 Trillion Market, $38 Billion Tokenized

Jeremy Au: One thing I'm curious about is the market size of this opportunity. In finance, there are a lot of fancy dinners, and everyone has a nice corporate headquarters. But how do you define the market size for real-world assets, adding in other angles like tokenization and agentic finance? How do you define the market Conduit is going after?

Richard Schroder: The first place to look is the existing RWA, and that's around 20 trillion. What's tokenized to date is around 38 billion.

Jeremy Au: So that's really, really small.

Richard Schroder: By any measure, that's where we're at.

Jeremy Au: The beginning of...

Richard Schroder: The journey, yeah. We're in the first chapter. There's a long way to go and a lot of assets to tokenize.

Jeremy Au: When you think about that, how fast do you think it'll get tokenized?

Richard Schroder: That's the $20 trillion question: how fast will this play out? It's the speed at which we can show the proof points for why being in tokenized form creates additional utility. Then there's the speed of what's happening in AI, and whether my view holds that agents will prefer to interact with stablecoins, tokenized RWA and crypto rather than with the traditional fiat banking system.

How Conduit Is Different

Jeremy Au: How is your team approaching this differently from the other players going after this market?

Richard Schroder: We're really fortunate to have a combined Web3 and AI capability that can also deliver in the institutional-grade setting I've described. I think that's key for us on the tech side. The other side is that we're really well licensed for this region, either directly or through partners, in Australia, Singapore and the BVI, where we're a regulated fund. Those are the two key areas. The other thing that makes us different is being firmly based in APAC and servicing APAC clients. A lot of this has been pretty US-centric to date, and we call APAC home.

Jeremy Au: Fantastic. There are a lot of clients in APAC that need your help and support, both in thinking through what to tokenize and how to get into this category, and in picking the right funds and choosing the right opportunities to go after.

Advice for Fund Managers

Jeremy Au: For any fund managers thinking about tokenizing their fund and taking the steps you have, what advice would you give them?

Richard Schroder: The most important thing is to make a start, if you've formed the view and done the research. Larry Fink says that 100% of all asset management is going on chain. At the moment, we're not even at 1%, so the answer is somewhere in between. But it's definitely only going to increase from here. So make that start.

What we've tried to put in place is this: work with someone like Conduit Digital Holdings, ideally with Conduit Digital Holdings, so that you can start with a fund-of-funds approach. You go out under our licenses, we tokenize it, we buy the underlying, and then we observe together. It's the lowest-risk, lowest-cost way of doing it. Observe together: is it establishing additional utility? Is there product-market fit? Are proof points coming through as to why to tokenize? Then evolve to tokenizing your own funds in earnest.

That's the tokenization-as-a-service and agentic services side of our business, where we're a technology services provider enabling those fund managers to take a tranche of their existing funds, or a new greenfield fund, to market. Ultimately, if it goes as I think it will, all the new funds coming through will go out in tokenized form from day one.

Jeremy Au: Amazing.

A Brave Moment: The First Stablecoin Trade

Jeremy Au: Another question I want to ask you: do you have a personal or professional story of a time you've been brave?

Richard Schroder: I remember very vividly doing the stablecoin transaction. It was with one of the bank's oldest-standing clients, the Smorgon Group, a client for over 100 years. I was more focused on disaster recovery, on what would happen if something went wrong. Of course, nothing did go wrong with the 30 million stablecoin trade, but it felt a bit brave at the time.

But perhaps, more broadly, being brave was evolving from a product coverage banker into more of a technology focus, following these exponential technologies.

The Team Behind Conduit

Jeremy Au: Let's talk about your team. Obviously, nobody can know everything, and nobody knows much about agent finance because it's such a new field. What's the team you've built at Conduit?

Richard Schroder: I've been really fortunate in the founding partners I have in Conduit Group and Cubewire. Cubewire, on the tech side, is genuinely impressive. We've got dedicated infrastructure in our offices for large language models, working with Nvidia chips and putting them in place, and that really helps us. We don't just access models from other providers. We also have an in-house capability on the AI side, and that's a point of difference, to be able to combine it directly with the Web3 side of things.

Then in Conduit Group, I'm very fortunate to have a well-licensed, very experienced finance team across all facets, and good guys to boot. It's that combination we always wanted for the tokenization of financial instruments: licenses and tech capability. Even though we mostly provide, and seek to provide, a tokenization-as-a-service and agentic services proposition, we think we're better placed if we can work right alongside our customers and walk in their shoes, particularly around what it is to be a regulated finance company in our own right.

Why Build This Now

Jeremy Au: I'm curious, because you could just retire, or you could have stayed in banking. And here you are, building at the frontier of finance, agents and tokenization. What's driving you to push with this?

Richard Schroder: It was the right time for me, in terms of family and personal position. And I thought that I, with my team around me, have as much right to do it and as much capability to do it as anyone. So if we can do it, should we do it? Yes, we should, because there's a lot of room for improvement in the provision of financial services: access, efficiencies, getting rid of a few intermediaries, making them high quality.

There's also the issue that about every seven years there's a banking crisis. We can evolve to a better provision of financial services for the customers we serve. I fundamentally believe there's that gap, that delta, that can be achieved, and we're going after it.

Jeremy Au: Thanks, Richard, for these insights. For those who want to know more about you and your work at Conduit, how can they find or follow you?

Richard Schroder: There's the Conduit Digital website. There's also the broader part of the business through our shareholding, Conduit Group, and you can find us there. They have a very capable broader sales team to support. And of course Telegram and LinkedIn. Find me there. I'd love to connect with you all on LinkedIn, for those who have found this discussion interesting.

Jeremy Au: Thank you so much for sharing. You're really a pioneer in getting this all set up, and I'm so excited to see what the next chapter brings.

Richard Schroder: Thanks so much, Jeremy.