Knova's Natalya Thakur on Canton, the 90/10 Split, and the Palantir Play for Institutions

How is everyone doing? We're we're great. Like, we've grown up, I think, so much since the last time we we talked with you. It's like going going live. We were talking about it. It's a it's a little bit smoother. We're finding the transitions. V two of our studio is about to launch, so I think we're we're I saw the pictures. A little bit.

So That's awesome. Well, exciting time in digital assets. I was gonna say, like, wow. We're really moving away from experimentation into enterprise deployment, like, Clarity passed.

I don't know if you guys saw the news out of Canton this morning. I mean, there's just been so much. Yep. So it's been really crazy. So so what so from your from your vantage point, we were actually just talking about this, I think, twenty twenty minutes ago before you hopped on. You opened your time line this morning. You see that news. What's the first thought that kinda goes through your head?

Well, I mean, first of all, I'm just not surprised. Canton's, you know, been working with the same group of players and people for a very long time. And I think what we're seeing is just, like, people doubling down more and more. And I think more if if we take this out of the context of, like, the news this morning, but more just like what we're seeing with clients, New strategies, new technologies, new emerging trends across digital assets are only, you know, exponentially growing. Right?

We're gonna see that continue. I don't think that's gonna slow down. And I think what we've been really helping clients do well is how do you deploy and try and test new strategies very quickly without having to have, like, a new twelve, eighteen month infrastructure build every single time. Right? Because we work with mostly different types of traditional financial institutions like asset managers, banks, payment networks.

And I think in a world where things are already moving so fast in the news, that means people also want to try and experiment different things that are gonna bring in revenue for their business or kind of future proof their business. Right? So we've been doing a lot of work with them where we're saying, hey. We're an on prem solution. We work alongside your existing back office systems. We're not doing a rip and replace or anything. Like, we're gonna make this easy for you.

You can work with any vendor if you wanna work with vaults or if you wanna work with different stablecoin providers. We don't care. You know, we're agnostic to that. Sort of trying to become like this, like, Palantir for financial institutions adopting digital assets. And I'll actually give you a couple of examples, like, with emerging trends. Like, even in the last month, what we've heard a lot about from clients, like, I think, you know, two months ago, everyone heard about, like, stables, tokenized deposits.

That still continues to be a thing. But I'll give a couple of examples from this month that are Canton and actually Vaults. So a couple can like, with Canton, we actually launched our global strategic partnership with them earlier this year. And so with Canton, there are a lot of people that are looking to not only just, like, connect to Canton. Connecting to Canton is, you know, the easy part, but it's like, how do I actually operationalize Canton against my existing systems? Right?

There's a lot of stuff on the client side that needs to be solved for that Canton is not necessarily doing for them. Right? So that might be DAML development, their smart contracts, party provisioning, key management, integration back into their existing systems for auditability and reconciliation because a lot of people on Canton are highly regulated. And so for most institutions, they're like, you know, I'm a super validator. I got to reach my milestones or I want to become a super validator.

I want to be earning rewards on Canton, but it's gonna take me nine, twelve, eighteen months before I actually get a production grade transaction moving end to end. Right? And no one wants to wait for that, especially when Canton is paying them out in rewards. Right? So we've helped them really accelerate, like, their current engine operations teams to get those rewards and get access to that system and using that system within weeks versus, like, many, many long months.

And we actually partner with likes of, like, AWS and DRW and other very large institutions that support the work that we do with our clients. So I would say that's one bucket. Another emerging trend that we saw was actually with vaults. I'm sure you guys have seen vaults are, like, all the rage right now. I literally was just messaging, like, love vaults. Let's talk more about vaults. That's been a Yeah. Great topic. So how I see this is really about, like, okay. The future is looking a lot more programmable. Right?

So you can imagine things like automated treasury strategies or tokenized collateral movement or yield products. That's, like, the most, like, easy thing that I think people grasp onto or just, a more dynamic movement of capital. But from our perspective, you again, going back to this thing of, like, how are we helping enable people with these strategies? You have the vault, but that's not the hard part. Right? The hard part of the vault is everything around it.

So, like, a lot of institutions don't realize when they're, like, getting excited about vaults is that they internally are gonna need to answer questions like, you know, what can initiate who can initiate movement with the vault? What approvals are required? How does compliance interact with the workflow? How do you report and reconcile the activity in the vault with your internal systems? And so that's actually not a vault problem. It's an infrastructure problem.

So I'm kind of thinking the application layer of finance will only actually scale if the institutions are operationally sound from the infra perspective. So a great analog I use with our clients is the cloud infrastructure. Companies don't build their own data centers every time they launch a new application on the cloud. Right? They build differentiated products on top of reliable infrastructure. So with that, I we're really trying to say, like, great. You wanna work with one vault, three vaults.

You have all these strategies you wanna deploy. That's great. We're gonna help you on the operation side of it to make sure that you're answering all the right questions to make sure your vaults work with your existing systems. It seems like there's been a huge transition of, like, one, getting started quickly,

like you're mentioning. Like, getting started is becoming getting which is, like, so key. Like, getting people off of zero is becoming less intimidating and getting getting something to market, but then also not having to orchestrate or build something from zero.

You can actually it seems like there's a sense of trust that's building with institutions from what you're seeing is, like, we're actually seeing a rising of what's possible, but also the trust that institutions have toward, yeah, getting started and using some of these tools.

Yeah. I think probably the biggest thing we come up against is, okay, you have a budget that's unlocked. People are really excited about a strategy. You know, someone in the c suite or head of BD is like, okay. Go build this. Like, this is what we're doing. And then you have a poor head of, like, operations or product holding the baton saying, okay. Like, how am I gonna do this? And it's the classic, like, build versus buy dilemma. Right? A lot of these institutions have a bias towards building.

And especially in the age of AI, people think even if they can't can't, people have this, like, false sense of, like, with AI, like, I can just build everything and anything. Right? So, like, we're just gonna even build even more in house than we were before. And I just think that that is going to be a pretty misleading misstep. I mean, we've already seen clients, like, have that have gone that way and within the span of, like, four or five months have come back and said, actually, like, we haven't really done much.

I think how I think about it is the ninety ten split. And I think this is sort of my framework for everyone moving forward. Because I agree. I think people should be building some stuff in house. Right? But it doesn't have to be the net of a 100. So when I think about the ninety ten split, I think the 90 is like every institution is gonna need certain set of capabilities.

So 90% for for enabling digital assets with your system is like asset normalization, right, working across different assets or ledger infrastructure or permissions or transaction workflows or connectivity. And so when we think about that or even auditing, for example. But every institution needs those capabilities. Then it's the 10% is where the firms are differentiated. Right? So what are their products, their client experiences, their risk modules, their proprietary workflows?

You want your best engineers working on that 10%, not on the 90%. Right? And so that's something I think that's super important. We actually, like we're looking at some stats that McKinsey had mentioned That analysis shows that institutions who actually utilize that ninety ten per in that 90% are utilizing modern modular infrastructure. They achieve 20 to 50% faster timelines to market. Right? Mhmm. And in digital assets, that's the game. Like, how fast can you get to, time to market?

And so I think the goal is for the 90% that's normalized, you should partner. You should get that 90% up and running. For the 10% that's specialized, your best engineers and your internal resources should be, like, maximizing on that 10%.

And so that's how I would think about, like, where you're choosing to build versus where you're choosing to buy. I I wanted to circle back because you said some oh, Drew. If you have something on this, go here, then I'm gonna circle back actually to what you were talking about, rewards. But, Drew, do you have something Go ahead. Here Go right now? No. So how you said that they're are they coming to you kind of already aware of how this reward ecosystem works? Did they see something online? They're like, wait.

I you tell me I can build it this way and and make that that much. You know, we've briefly touched on that in some of our previous conversations, but maybe for people that don't know how that rewards ecosystem actually works with being a validator

and being on the network? Can you kinda break those down and give us some insight into what those conversations have looked like? Yeah. So there are different economics. Like, obviously, if you're running a validator, you're direct access to rewards. You can also not run a validator and still be engaging with Canton applications through other providers and still be getting cuts of those rewards as well. Right? There's there's different ways to do it. Right?

That's just like people hacking the business model of making it how can we all kind of, like, share a part of that pool? I think there's probably, like, different buckets of people who understand how they need to work with and implement rewards. You have your first bucket, which I kinda call your zero to one people. They're mostly in advisory mode. Right? You're kind of they don't under they don't have a digital asset strategy today.

They don't really understand Canton, but Canton is the, you know, financial institution safe route in their heads. They need to figure out how to even get up and running. Like, what's the simplest thing they can do on Canton to earn rewards, get in the game, start doing it in a way that's like very much they can sell it internally. Right? Because they have to get approvals internal. Right? It's all about how can they champion it in a way that doesn't get them fired. Right?

If they're gonna do this, how do I make sure I don't lose my job? How do I make sure I get promoted because of it? Right? That's number one. Number two is the group of super validators or validators. So these are people who are very comfortable with the economics of the rewards because they're building on the system. They are joined as a validator, they understand economics of it. And that's just for normal validators. For super validators, it's even more intense on getting time to market on those rewards.

Supervalidators have set milestones that they need to hit within a certain period of time on the network. And to hit those milestones means their payout is also much higher from a rewards perspective. But if you need to hit those milestones very quickly within a set amount of time that's based off by the network, that means whatever you need to do to operationalize, Canton needs to be built in house. Right?

We're talking to a client right now where they have to build a bunch of stuff to operationalize the milestones that they've set with the network. They need to they're a large institution. That could take them multiple months that cut into their nine month timeline that they have. Right? What kind of milestones are we talking about here? Like, do you have any examples to help people think through, like, what an example of a milestone might be? Totally. So every milestone per supervalidator is different.

It's based on their business. So So that and they work with Canton to determine that? Yes. You work exactly. You work directly with the foundation. It goes through a committee review process. You work directly with foundation. You basically say, like, here's my business. Here are the milestones that, you know, are relevant to the work that we do, like stablecoin settlement, for example. And basically, they get approvals for each of those milestones and then how much weight each of those milestones have.

So it's a pretty It's sort of similar to, like, an early employee having, like, stock grants or something. Yeah. It's but but it's a pretty, I would say, manual process in that you kinda have to sit down, come up with the milestones, get them approved. It's not like an easy thing versus when you're just a normal validator, you're just building whatever. And, like, as long as you put transactions on the systems, you get rewarded.

Super validators, it's like intentional work that's built into these milestones that have to get approved. And then, you know, once they start, you know, pushing transactions through those systems and get reaching those milestones, their payout is monumentally or exponentially higher. And then I would say beyond the validator, super validators, you have a net ecosystem of very large players that are investors in the space that, hey. They're just here to see like, we wanna see capital markets go on chain.

We're here to support. A lot of the cap markets work on chain requires multiple stakeholders at the table. All those stakeholders are the largest in the world, right, like the citadels of the world, the trade webs, and they're also kinda plugged in. And what their their goal is is to figure out what tokenization flows or operations or use cases are gonna be the best ones, and then how can they become, like, the market maker in the system, or how can they become, like, the default application provider in the system?

So they're kinda playing the long game because those stakeholders service every single part of a capital markets trade. So you have 52 super validators,

like, we're showing at least at this point. And 700 total. Total. So Wow. The 52 are kind of in this, it seems like, special special group. Is it also more lucrative, I'd imagine, but more involved, it sounds like, with them?

Super validators, highly more lucrative, highly more involved. Yes. And on purpose. Right? I mean, like, you're you're choosing to do that. And actually, I will say for those that are just in the normal validator bucket, there are a lot of, like, components of the ecosystem that try and help support you to put up a super validator application. So, like, that's something that DRW actually does, you know, with, you know, select validators.

Like, they will help you put together that super validator application, help you go through the process, and make sure that you're kind of handheld through it. What do you think people most

or what what do you think people miss the mark on right now with Canton? Right? Because they kinda put them into this, like, mental model. It's like the privacy institutional network. And I know we've talked with Eric and Yuval, and they've kinda, like they kinda go, yeah. But, like, there's there's a lot more to understand. Like, how what is that what your mental model is for thinking about Canton in the digital asset ecosystem, or do you see it, you know, evolving?

I don't know. I mean, I would honestly leave it up to that team to probably talk to this. Right? I'm not like I'm not Yeah. I'm not digital asset. I would say that my sense is probably, like, yes. I think the default wording to kind of say privacy enabled, financial institution foolproof, like, that's easy to just get people to understand how it might different from other l ones.

But I think the point is that there's a ton of, like, intricate work that happens in capital markets flows across products where their ledger has to be nuanced enough to support those things. And those are probably the things that they're kind of like, it's a lot more than that. Right? But it's hard to just kind of explain off the bat if you're not familiar with how those markets work. But, yeah, I don't know if that answers your question.

No. No. That's great. Maybe maybe the better way to frame it would be, it sounds like Nova and and you personally, given your background, are in the position of, like, playing almost like a quarterback role or, like, a consultative role with with those that you work with.

Help us understand, like, where Nova plays in this conversation and how you guys are positioned in and around things like Canton and how the clients that you're interfacing with and the partners you have, like, what is the the core the evolved like, here's how we're providing value on behalf of Nova. Yeah. So from our perspective,

like, if you're a good partner, like, a lot of people have questions that's early on. Right? So, like, I'm not saying our business is not advisory, but, of course, like, we're gonna sit and advise our clients if they have questions. Right? Like, there's, like, all the questions you're asking today, we've heard those a million times over with other people. But I would say that from a tech perspective, we're really the this toolkit that helps them operationalize what they need to get done on Canton. Right?

So they might have a vision of, like, what do I wanna do on Canton? What is my business really good at? Like, where where is it lucrative for me to, like, place a bet because, like, this is how we make money? But operationalizing Canton against your existing systems is hard. Right? Like, there are so many different digital asset products. Everyone wants to, you know, take a bite at the apple or do different things with different partners. That's great. You pick your set of vendors.

In this case, we're talking about Canton. But there's a bunch of stuff you need to do internally on your side to enable all your other systems to sort of, like, work with Canton for you to reach those milestones. Right? Canton can't be just used in a silo. There's other, you know, cash accounts or wallets or custodians. There's so many other players outside of the Canton ecosystem for a financial institution to make it work within their four walls. And so what we do is we really help players.

Photo for this kind of talk. I guess, like, in this case, like, one of these folks would be Canton, but, like, they're basically saying that, like, there's other sports involved. Right? There's no institution that's, like, only using Canton and doesn't have access to any other fiat rails or any other, like, providers or wallets or banks or custodians.

It's just there are people working in financial institutions and people who are in the financial market system work across a multitude of external systems, Canton being one of them. With Canton on specific, we will help come and make it super seamless for them to run their operations on Canton. Right? So now they can call a single REST API call or now they can do a very simple operation, but they can track everything. Right? So we can help them with DAML and smart contracting development for them.

We can help them with party provisioning. We can help them with key management. We can help them with integrating some of the existing systems to the Canton ecosystem for, like, auditability and reconciliation. Right? There's a bunch of things of how data is stored, kept on Canton, how that kinda maps back to their existing systems. Canton today doesn't have the flexibility, and they wouldn't want to be a dev shop essentially for every single one of these clients. Right?

So what we've done is created a normalized toolkit layer that makes it really easy to use Canton so you feel more comfortable doing more strategies on Canton. You feel comfortable putting more flows through Canton because you don't have to sit and, like, hire a massive team, train them up on, you know, certain ways to code and develop against smart contracts and then, like, figure out, okay, what is, like, safe enough to do?

You know, a lot of the early Canton, how do I say this, POCs or pilots, they were doing, like, a thousand dollars. Mhmm. Right? Canton doesn't want people to just be pushing a thousand dollars for, like, a repo transaction. They want people to put real money on the systems. But to do that, people need to get comfortable with how they're, on a back office perspective,

mapping everything and tracking everything. Right? Mhmm. So that's that that toolkit help is really we know where we kinda come in. It seems like there's a huge, like, one value add in here's all this fragmentation that exists for these people. Right. These companies who have all these accounts, all of these chains, all of these integrations. Like, the fragmentation problem seems to be a huge one that Nova is solving.

What are these other, like, key issues that you're finding is, like, a really big value add that you're bringing to the table here?

Yeah. I think fragmentation, solving that is really around, like, the readability view this visibility piece, right, the orchestration part. And I know a lot of people use the word orchestrating and mean different things by it. So just to kinda set that as a standard. But I think the second is really operations.

Being able to run operations against new systems, have that vendor flexibility is extremely important because today you might be using one custodian in six months that might be different, and you don't wanna have to keep rebuilding your internal workflows and infrastructure to support new strategies and vendors. I think the way that we position ourselves is operationally, you can start today with a certain strategy. In five months, you might wanna add a new strategy or change that.

Functionally, your base layer with us is, like, set. Right? You can you can scale across multiple strategies using our toolkit without having to start a new siloed infrastructure build for staking, a new siloed infrastructure build for your vault business, a new siloed infrastructure build for whatever you're doing on Canton. That is also increasing fragmentation. So if you have one toolkit that can kind of say, hey.

We can kind of support you through scaling across these different digital asset strategies, and sometimes those strategies might even intersect. Right? You can imagine a world where, like, your stablecoin and your vault strategies start to intersect. You wanna be able to have that operational unified operational layer that allows you to do that if that becomes the case. What is the existing

operational layer? Because the source of truth, it seems like in this mindset is, like, Nova will be the partner to stitch these things together, be the source of truth, make life easy, and it kinda goes back to what you were saying in the beginning of, like, testing new programs, piloting piloting new things. If you have an organized core, then you can go and test and, like, maybe try things more easily where it doesn't actually increase fragmentation issues. What is what is that, like, replacing in your mind?

It's I'm imagining there might be, like, incumbent source of truth systems that are being used today that don't work. Like, what what does that look like?

Well, I think what what we are sort of repainting is today, if someone sometimes if they build a pilot or POC, that's in a closed environment, and that tech does not work for production. Wow. What a waste of time. Right? You got the check mark on the POC, but, like, you're never gonna be able to use that. Or they build something that's for stablecoins, but that won't actually extend to other, you know, other systems. So what you're seeing is, like, a fragmentation of projects that are not stitched together.

And usually what happens is that there's a digital asset org that owns all these, like, fragmented products or projects stitched together or are not stitched together, and then you have, like, the traditional asset side of the house. Now we are not trying to replace those core systems. The core systems for the existing workflows and mainframes that they have, those stay the same.

What we say is we sit adjacent to those systems, And we say your core systems and your mainframes, you have current workflows that work for your traditional side of the house. As you're adding new digital asset strategies, they all have different workflows and new workflow systems. Today, those two workflows do not talk. Digital asset workflows across various projects and your mainframe workflows don't talk to each other.

So we help neutralize those workflows into a common standard language, if you will, to just for simplicity's sake. So that's really what it's about. It's not replacing the core. It's sitting adjacent to those core systems, but it's translate translating core workflows to digital asset workflows and making sure that they can talk to each other. Yeah. That's amazing. And so you're also working with maybe some of those traditional and and core workflows. It's not a rip rip and replace.

It's making Sometimes we all do things easy. Honestly, I think a lot of us in industry would love for that to be the case, but that would just it's a 0% chance that that's happening. Right? Like, there is no large institution. I mean, let's put aside banks. Even, like, more tech forward institutions like some of the hedge funds or asset managers, no one's ripping and replacing whatever core system they have. And we also don't wanna be in the business of being a core system like that. Right?

Like, we are a control plane. We are a operational layer that sits adjacent to those core systems. Our goal is to help that translation happen so that operations can run from a common set of, you know, common language, essentially.

Huge. It's always great having you on, Nat. And I love I love your thoughts on what's happening in the news and just kind of getting your pulse on stuff like the Canton Rays and that ecosystem. It's helpful it's helpful to, like, have someone like you who's, like, you know, in the foxhole, talking with people, building the business, also have commentary on, like, you know, what you see on the on X or whatever. It's like, you know, that for us, it's always helpful to have people be able to bring color to that.

So really, I know you I don't think you were expecting to, like, go that deep into that, but appreciate your x fun. Yeah. I know. Of course.

I'm like, digital asset, like, it's like digital assets gossip for breakfast. Yeah. I love it. I'm like, should we do a soft set we can do a gossip,

like, side thing, and I can say that. We need, like, a gossip hour, like, segment where it's like, hey. Let's just, like, throw aside yeah. We we can talk about our products and our companies all the in the news today? On in the news space right now. What's the tea? It's funny. Stable tea. Whatever. There we go. Stable tea that. Honestly,

please, I'm raising my hand now. I will leave that segment for you guys. Hilarious.

Hilarious. Would you might get a few DMs from Drew later today? There you go. There's a lot of tea. There's lot of tea. So There is. I have I have my teacup ready to go.

Love it. It's a little bit big for a teacup. But, Nat, thanks for the time. We'll have you back. Of course. Friend of the show. Hope to see you in person soon. Will you be in Mexico City by chance at all next week? I will not, but I will be at PZF in Zurich in a few weeks. So Cool. 0.0 Forum. Lots of TradFi's heading that way. So Love it. Love it. Alright. Great to see you. Have good one.