Adam Cai / Virgo

How Wallet V is letting AI agents trade while keeping your keys

InterviewNovember 10, 2026

In this episode

Adam Cai, founder and Group CEO of Virgo Group, is back on Crypto Coin Show to talk about Wallet V, the self-custody mobile wallet where AI agents trade for you while the keys stay on your phone. We get into why he moved from running the VirgoCX exchange to building a DeFi super app, what Wallet V's benchmark data shows about which AI models are winning, and why AI is a trading buddy rather than a shortcut to becoming a millionaire.

We also cover the Phase 3 campaign, where users can earn up to 14,000 USDC for bringing top Hyperliquid and AI traders onto Wallet V, how VPoints lead into the airdrop and the token launch planned for early 2027, the risk controls that stop an agent cold at a 20% loss, and how to get started with as little as $100.

Key takeaways
  • Wallet V is a self-custody mobile wallet where AI agents can trade while users retain private keys and full control of funds.
  • AI agents have built-in risk controls that stop trading automatically if losses reach 20% to protect user accounts.
  • Wallet V integrates native DeFi protocols like Hyperliquid and Aster Swap with gasless transactions to reduce entry barriers for users.
  • AI agent trading is positioned as a tool to help users execute their trading vision, not a guaranteed path to profits in complex markets.
  • Token launch is planned for early 2027 with Phase 3 campaign offering up to 14,000 USDC rewards for recruiting top traders.

Chapters

Read the interview

Wallet V's Adam Cai: AI agents can trade your funds but never hold your keys

Adam Cai, founder and group CEO of Virgo Group, the team behind the self-custody Wallet V, talks about AI agent trading, the limits of today's models and the safeguards that keep users' keys in their own hands. He also covers the early-user rewards campaign ahead of the token launch, which was described as planned for early next year. The conversation was recorded in November 2026.

From exchange to AI-first wallet

Ashton: What is Wallet V today, and who is it built for? 01:13

Adam Cai: We're seeing a strong trend of users wanting to control their own keys. They also want a DeFi experience that is native, user-friendly and simple to understand, and many want it on mobile. So the first reason we built Wallet V was to offer that native DeFi experience, a kind of DeFi super app. Users control their own keys and get easy access to things they love, like Hyperliquid, Polymarket and Aster. We did a lot of homework to bring top-quality protocols into the app, so users don't have to work out which is best. The entry barrier to DeFi, and to trading in general, is very high. Users don't understand gas fees, or what USDC means on different chains. So we built features like gasless transactions. We cover a certain number of transactions, and swaps over $5 are free of gas. AI helps on several fronts too, from a chatbot that answers questions, to technical analysis of charts, to autopilot trading connected with Hyperliquid and Aster, using either prompt-based models or an AI framework.

Ashton: You started with Virgo CX, a traditional exchange, and have moved to self-custody and AI. Why? 05:00

Adam Cai: I'm always trying to bring a product that makes an impact on the industry. We started Virgo CX as a Canadian trading platform on a fiat ramp, and in 2019 people needed that, because they had to be able to buy and sell crypto. But looking at 2024 and 2025, I saw that the infrastructure for crypto on-ramps and off-ramps was very mature. There isn't much innovation left to do. It's really about compliance and banking access. I wanted to do something more exciting that brings substantial value to users. So I asked what the future of a mobile wallet is. A crypto wallet won't just be storage, like a hardware wallet. It will be a super terminal for global users to access digital assets: not only crypto but also stocks, commodities, yield farming, everything about finance, following the Hyperliquid vision. A wallet is very flexible and compatible with how DeFi and tokens are designed, so the speed of iteration is a power I have never seen with a central exchange. We can connect directly to excellent protocols through smart contract APIs, without even talking to them, and pick the best one for users.

Read the full interview →

Clips from this interview

By the numbers

250,000+Wallets created in past two months
$1B+Trading volume achieved to date
5,000AI agents run for performance benchmarking
5%AI trading percentage of total volume
20%Maximum loss threshold before AI stops

Transcript

Read the full transcript 5,478 words, auto-generated and lightly edited

Should I trust an AI agent with my funds?

The AI does not have private key. Okay, the private key is still owned by the user. All the AI can do is trading. They cannot move your funds anywhere else if your total loss hits 20%. So AI will just stop. Nobody can touch your accounts other than yourself. I'm Ashton Addison from the Crypto Coin Show and today we have back with us Adam

Kai, the founder and CEO of Virgo Group, the team behind Wallet V, a self-custody wallet where AI agents trade for you while the keys are still in your hands. Their token launches early next year. And right now, early users are earning rewards ahead of the token launch in the wallet v mobile wallet. So today we're going to get into AI agents, what the

data says about who is actually winning and what's coming before the token launch and how AI is revolutionizing all of crypto and trading and much more. Adam, welcome back to the show.

Thank you Ashton for inviting me. It's always great to speak with you again and great to speak with audience and glad to share what's the latest with us. Yeah, so much has happened since our

last conversation at Consensus in Miami. Wallet V has been growing fast and I consider it to be the leading AI first mobile wallet in crypto. And I have a feeling like AI is going to take over all exchanges, wallets, everything. and just and all of work outside of crypto as well which we can get into but I'd love to start high level with the latest

on what is wallet v today and who is it being built for.

Sure. So W let's kind of starting with the I guess the reason you know why we started W project we're seeing a strong trend of you know user trying to control their own key and we're also seeing user trying to get a platform where the D5 experience is very native it's very you know user

friendly simple to understand we also would like a lot of users would love to have that experiences on the mobile app So initially the first reason for us to build WV is to provide that native D5 experis like a D5 super app kind of a feeling to the users so they can kind of control their own key they can get easy access to D5 they can use the

things they love like such as hyper liquid party market aster swap earming like we just built that natively and also we did a quite a bit of futuring trying to really kind of bring the best you know the top-notch D5 protocols to the app without you know giving user the hassle to figure out what's the best for them right so we did a quite a bit of homework to bring the

best you know quality ones to the app and then we're seeing that you know the problem for user to using DI and also even to just for trading in general speaking is the entry barrier for that is actually very high let me put that way

the user is not understanding what is you know gas fee the user doesn't understand you know USDC for a

different chance what does that mean user didn't understand about how to interact with E5 properly so we are making quite a bit of features including the gasless transactions so user can easily send receive the cryptos without paying the gas we cover a certain number of transactions

they can do the swap as long as it's more than $5 it's free of gas they can

departed files to poly market support and hyperl liquid without g a gas fee needed. So we actually made quite a bit optimization to load edge barrier and also we introduced the idea of AI. So AI is helping our user in wallet B in multiple fronts like starting as small as chat box right to telling them what's available in the wallet helping

them answer different type of questions that would be like some simple knowledge about what's crypto and what's wallet B and what's the feature and also can go to analysis of the Kline right technical analysis and also can go to the kind of the real usage of the automatic autopilot trading connecting with the hyper liquid masteraster using either

be like the promp based model or could be like AI framework when it comes to the crown trading. So we do have a full-fledged of AI features building w trying to really help the users to get easy access for one and two is trying to get them a long-term survival rate within the defi network and trying to really help them to kind of become active users within the digital asset

space. there's a lot of functions to cover especially as the industry expands out in different directions and adds AI and tokenization and more types of assets. Originally when we spoke many years now we've been on the show 8 n years the team was building Virgo CX as a traditional exchange and then now you've transitioned over to the wallet

and self custody and AI. What made you go in that direction?

Right. I think the main reason for me is you know from a high level I'm always trying to bring a product that can make impact to the industry right we start off as a verbal CX as a Canadian you know trading platform on a puff ramp so that's kind of I would say the early taste of transitioning my kind of career from

trify to crypto and at the time of 2019 people do need the services right because people need to be able to buy to sell crypto but while we're looking at you know 2024 25 I seen the in structure for crypto onrap and off is very mature right there's not that much of innovation you can do it's really about compliance about banking access so I feel like you know after five years in

the kind of the I would call it kind of trify combination of trify to crypto I want to do something that is more exciting I want to do something that's bring substantial values to our users. Then we started kind of think of what can we do. So WY V naturally comes to the picture where I kind of envision the future kind of a digital asset world the future you know what is

that actually the future of a mobile wallet like what does crypto wallet actually mean right is actually just like a hardware wallet is just for storage my answer to that is no crypto wallet in the future will be a super terminal for global users to get access to digital assets right not only crypto but also onchain AWS, stocks, commodities, yo farming, a bunch of

things, right? Everything about finance, right? Kind of a coding the hyperlquid vision. So, wallet is actually an important infrastructure for that. And because wallet is so I would say how do you call it's so much flexible and so much compatible with how DeFi is designed, how tokens designed. So the ability to iterate for the innovation within a wallet is the speed. It is a

power I have never seen with a central exchange, right? We can really stand on top of the real excellent talents, the real excellent protocols they have built and we don't even have to talk with them. We can directly connect with their protocols through smart contract APIs and we can pick the best one for users. So that interpretability, right, when we

talk about that, right, about you know that kind of ability is so fascinating and there's so much value we can generate from that. Yeah, it is very exciting and it's just like how the mobile phone you know it has a thousand apps and I feel and you have your phone in one pocket, you have your wallet in the other pocket that's just going to all

combine into one device and connect into every single financial service and application but inside of your own pocket. you don't have to go to the bank anymore, which which no one likes doing. But one one thing I am curious about with regards to AI agents, there's been a lot of talk throughout 2026 that AI agent trading is the future, but we're very

early and we haven't really seen that. It's still an early experimentation phase. H how big of a bet is this with wallet V that this is the future and is being wrong or being too early the same as being wrong? That's a really good question Ashton. I do love that question. I have been asked by many many different partners about that. So I think to start with I

think there's no doubt AI will help us in a tremendous way for many different aspects and the trading is just one of them and I think you are 100% right the AI agent trading is at the very early days actually very early and I think really when it comes to trading like when it comes to financial market this is I think there's no doubt the capital markets the stock market the crypto

market is the most difficult and complex markets out there, right? So there's no fixed set of winning rules how everyone can make money, right? Because if we can, we don't have to do anything else. It's just start trading. So I think the rules is very complex and there's actually no repeated success. So I think the AI agent trading is by no means a guaranteeing of winning for the

users. rather I think it is trying to help users to trying to realize a single user's vision about what's their trading style and help them to understand what's their risk management and help them to get away with the emotional trading activities which get get then wiped out in the next day no matter how much money make so AI agent in the current

days I think it's more kind of like a assistant that is based on a particular user's vision, a training plan and trying to help them to articulate what that makes in a very predicted manner or I would say a preset rules right based on how you interact with AI. So having said that if a user is able to generate offer over time with

the agent the goal is trying to make that success to be more controlled right and more kind of like a repeated success and if the user really bad at trading but they still want to make money

AI is there ideally trying to help them to kind of understanding okay whether your trading thing makes sense. has and try to understand what's the logic behind trade right be you know either

for technical analysis or for kind of trend following you know for kind of different indicators and also in the meantime try to help them to control the risks so they can learn from what they have you know made a mistake on they don't have to be wiped out and they kind of just say okay crypto scan and I left the industry so AI is by no means of kind of like a legendary piece where

you suddenly just become familiar with AI rather AI is a buddy and hopefully you can grab that power to kind of revolutionize not revolution that word is kind of too too fancy I would say help you to kind of grow yourself along with the understand of AI and truly understand who you are how your trading should be what's your return target what's a risk target and then in the end

you can let AI to do what it has to be done based on your demand. So that's how we kind of design AI and trying to really focus on that area. It's actually very difficult because there's a lot of hallucinations. There's a lot of technical details why sometimes air can can be messed up. But we're trying to help you to chos. That's a you mentioned about the

hallucinations there. There's many different AI models. The frontier models, Chinese AI models, open- source AI models. They all have different levels of knowledge and of hallucination. How do you determine the best AI models to use for Wallet V and for trading,

right? I think actually the way we kind of envision this is not how we decided is we would actually give all

the models to the users and for them to choose. We have the free models which we subsidize users cost because these ones are cheap. User can just directly choose those free ones to have a taste. We also have the advanced models right basically the latest release from TRGBT from you know codecs you know from different versions and they can just

literally just choose those ones to load their AR credits trying to consume it. So really we actually published right a little performance report based on the five to six,000 agents we have so far run and to show you that okay you know what based on our limited data sample and based on a similar prompt provided to the AI and these are the performance

For different models

and I have to say that you know the data sample is not that big because it's only 5,000 however it does give you some early sign

of which model can perform better under what market conditions and each model does have its a little bit different strengths and a wickedness right and I think that's the fascinating part right

it's not one model this wing all it's just not the user have to understand which model does what and which strength they have and they should be actually trying to pick the right model the right prompt with the right setting for different mark conditions right so I think that's where we're seeing Our users is really debating. We have seen users speak highly about agents. Some

people making I don't know like 100% return in three days. And I have always seen we have also seen people really say bad words right. Oh this a model just sucks. I lose all the money. Not all the money but lose a good good amount of money. And it's a lot of debate but in the end of days it's not what we provide. It's what will empower users to make their own choice. I think that's

how V should be becoming. Mhm. Yeah. The benchmark report showing the different AI models and over half being profitable over 50% is a good mark in the early stages of this. But as the models evolve, it's seemingly they're coming out with new models every month and they're getting smarter and I don't expect them to be trading for you with 100% win

rate. That's probably impossible. But how will you adapt wallet v's AI strategies or the education for the users as the wallet model the AI models improve

right I think you know this is actually a very like a very frequent topic that our internal team was discussing on a daily basis so the AI module which comes to the large model the underlying

ARM so that one is always constantly evolving right but what we are building is like what agent we can provide to users what hardness we're doing how we can improve the memories and how we can kind of leverage AI to produce a framework for users to capitalize on right so there's a lot of a lot of combination of you know what we can do

and you know using what tool And like with the limited resources we have like I think AI is also like a huge universe right if we give user all the choice they choose they just kind of get lost they don't understand which one to choose and they don't have time to learn and I think our value proposition is okay you know we understand this is universe how can we make it so simple

we understand you know the user base that we're trying to attract and know what goal they want to achieve obviously everybody want to make money That part I understand. So making money is a very difficult topic, right? It's not easy. We have to say that. But how do you make money? What's a path to lead you to make money? We have to really kind of break

down to understand each path and which particular point pain points we're solving. So this is actually a very very complex topic which really to be honest with you, we don't have the full answer. If we do, we're probably really like a billionaire. like we don't have the full answer for that but we're trying to solve like certain steps along the way for people how can they increase their

winning ratio and how can they actually more importantly is to have the ability to say okay you know what next time if I work with w this particular feature I can have a higher chance of winning if we can actually make that happen we already provide substantial value to our users

definitely anything that can increase your chances even marginally your win

percentage rate is valuable tools in your pocket.

Totally.

And now talk to me about the early traders being rewarded for getting on and doing AI Asian trading and being a part of the benchmark report. And this is leading up to the launch of a wallet v token and right now you just started phase three. So can you talk about what

those traders are actually doing and how they're being incentivized to do so?

Of course. Yeah, we actually have launched two campus already. The first two camp focusing on the you know kind of bringing more users to create wallet to try it out and the second one in you know refer people to come here to trade and earn the kind of commissions v points as well as

achieving daily tasks to be a little bit more active. So the third campaign is actually very interesting. So we're trying to kind of targeting the onturn Wales depending on their address. So for whoever can identify which address this guy is associ with and if they can invite to Wally V to make some trade they can earn up to 14,000 USDC and a quite some hands on of the V points.

So the goal is trying to let those entre wells influencers people influential people to know about what it be to understand what product we offer and we would encourage people to bring them over to WB to make us a little bit more impactful and of course all the other activities encouraging the daily act active you know activities in the you know water v and commission

sharing it's all there

so We're trying to basically you know after we have achieved more than 250,000 wers created in the past 2 months over a billion dollar trading volume now we want to actually make the impact a bit bigger because I think we already have a very nice like wallet very nice tool for people to explore and now it's just for people to know about this to try about

this so this is actually our main focus

definitely and you know I've been experimenting with AI agent trading with wallet v and some other platforms. And I'm curious on why this campaign is important to target whale traders and not just your everyday small small fund managers.

Yeah, for sure. I think you know the

crypto market is a lot about influence, right? to who who is a big whales in this space and wherever they make some recommendations people listen right and we just want to kind of get to a shortcut of people knowing what to be right we're trying to get the right people to backing us and we're trying to get into the right circle right without spending millions of dollars so

I think the campaign is particularly trying to say okay you know what

if you guys we know that okay you're already doing a lot of trading on I don't know hyperlquid you're doing a lot of trading on party market. Okay, try that in what V and this is exactly the same feeling, the same taste you are having with Hyperl looking aster and the party market this

like we're not building new decks. without building new prediction market.

Mhm. And I have a feeling about what it be. have a feeling about a self-custody mobile app that is so like smoothless for the kind of the people and once they actually try it that word of mouth is really going to spread

and we're hoping that we can have the Twitter backing we can have the

endorsement from those people that just in general increase legitimacy about you know what an increase of credential right trustworthiness it's very very important because nowadays even though actually the previous hack is really about hardware wallet Right? We're talking about ledger, we talk about code card like these are really hardware wallet. So people are now

starting kind of worrying people you know any wallet oh you know what you are you having enough readiness are going to be hacked no it's just about that kind of a you know question the concerns we also have to tackle.

Mhm. when you have AI agents trading on your behalf, but in this case, you're holding on to the keys, do you see risks of fund loss or mistrading? And what like

as you said, people are being cautious because of centralized exchange gets hacked, hardware wallet gets hacked, should I trust an AI agent with my funds?

Yeah, that's actually a very brilliant question. So first of all let me clarify one thing is within WB once a when AI start to trade for you the AI does not have private key okay the private key is still owned by the user

what AI has just an API key that is you know managed by hyperlquid aster so really all the AI can do is trading they cannot move your fonts anywhere else so that's the first thing I want to clarify so the second thing is okay let's say I habitably believe they can really mess up with the fonts if they get really stupid or they get kind hallucination.

So the way we kind of tackle this is we always require user to set up a hard risk limit. Okay. So no matter what you put in the prompt the parameter you're setting we do have a risk parameter to say okay let's assume that if your total loss hits 20%. No matter what high stop I'm not do any trading. So AI will just stop the whole system will stop. M

so I think we have built quite a few

different risk parameters within the system to monitor that

and the second thing is we want to make sure that AI thinking process is as transparent as possible. We don't want to have some kind of backhand you know like AI tried to kind of intentionally do bad things because now people think oh you know AI can do bad things right and we do want to make sure that

you know that process is very transparent so we do have an abstract generated for every single AI decision

they have to literally show in what if like what's the AI thinking process and why they actually click buy while they click and tell us what's the reason. Give me the abstract. And that's also giving people the kind of confidence level. Okay, you know

what? If they lose money, oh, it's because of this reason. Okay, I have to find two my technical indicators. If they're making money, it's because I read. So, every single decision is transparent.

Yeah. No, I think that's super important to have the reason and that can also help iron out any of the hallucinations that you were talking about with AI

agents. We want to know the reason, especially if anything with these frontier models, what's the reason that they took this decision with your money so that you can learn from it to not do that again if you if you happen to lose?

100%.

Mhm. And with the phase 3 campaign that's happening right now, the whale traders and the regular everyday

traders, they're earning V points inside for trading and also just identifying the whales and the top traders, whether it's their usernames or getting them on boarded. And so there's Vpoints, but there's also USDC, which is super nice. What do the Vpoints do inside of Wallet V?

Right. So viewpoint now is a royalty point system, right? So it's really like

every time where people is making a revenue generation event for us, we would actually grant users viewpoints. So in the future the viewpoints will be converted to you know another token that are going to be deeply integrated with w through a different set of company and it's trying to empower the w ecosystem. So the user now is super excited to earning the V points and

they're trying to in the future have some airdrop of the tokens.

So really what we're trying to advocate is positive activities revenue generating events within WV and trying to real do you real user activities trying out different functions we have brought to them and you know they will just become rewarded because they're early adopters. That's nice. And

with the phase 3 campaign and earning points, some of these mainet launches, they were having people earn points for like years and it felt like there was never an end to it and people get tired. Is there any tenative timeline for the end of the campaign and the token launch? Yeah, the first season of the end of campaign is actually coming right up soon. Like it's going to be like I

think by December 3rd, the first three months of campaign will be down and you will be eligible potentially for the first air drop. And we're going to actually launch you know second season, third season and fourth season. But each season probably will be only lasting for a quarter or so. So our campaign is very short in general because we do want to let people to see

The results and they're gonna do more and all that kind of having a really positive feedback loop.

Okay, I'm looking forward to it. And looking towards the future, this big bet on AI agent trading in 2 years from now, how big of a percentage do you expect, at least in web 3, AI agents to be trading more than people manually?

I think that's still a long way to go to

be honest. That's a far long way to go. I think if I would say like from our data I would say out of the you know the a billion dollar training model probably only 5% is from AI.

Mhm.

Like at the stage I'm I'm already calling this a pretty good success to be honest.

Mhm.

To start with right and by what time is going to be like the reflection point

where people can say oh can I just take my money like now I'm seeing most people just putting 100 bucks right up to like $1,000. Some people put the most the maximum I've seen is $50,000. That's already quite a bit, right? I'm not seeing anything more than that. So, people are learning. They're trying to figure out what that is. It's a lot a lot about trial and error.

ultimately, everybody want to have that, you know, ability to let AI to be your true money manager. But this is a very complex pathway to hit that because of the audience difficulty challenge were mentioned a bit above. So would you encourage people to try it out and to kind of really innovate along with us to learn along

with us and we are the first batch of innovators in this particular space. And who knows like you know I don't really know like because AI is developing so fast I think it's a long way maybe it's not I don't know.

Yeah we will see. But I think it's important to learn about it now at least. Test the waters. What's the best way to learn about the AI agent

trading in wallet V, try it out and learn about the V points and the USDC rewards campaign.

Now, actually, it's much easier, right? In the past, you know, when it comes to, you know, we have different shows, right? For Canadian exchange, you have to be Canadian users, unfortunately. But for WV, you can be anywhere else in the world apart from, you know, some

countries. whatever you can just download it from Google store, Apple store, search wallet v stand for rogo very easy just download that and creating wallet probably takes you 60 seconds because there's no KY required and yeah and then just move on a stable coin over and try our gasless experiences deposit 100 bucks maybe or if you

feel you want deposit less feel put it in and try different features. It's all DeFi based. If you don't like it, you can move out of money anytime. And the minimum for AI is 100 bucks because we do subsidize your AI token. If you put $1, it's just going to be too costly for us.

Mhm.

so 100 bucks I think everybody can afford most likely.

Obviously, this is not an investment

advice, but you know, try it out. try it out through different settings, try poly market, try different features we have and I'm sure once you started to use it, you will like it. You would say okay this is so actually similar to the centralized exchange experience we're getting at and why should I not put more money into it because and now

in that days even what V goes under we can never touch your phone as long as you have make sure you properly back up or recovery keys we offer two different versions you can put on your paper write down pick a banking watch or you can actually through the Google you know cloud Apple cloud you can do it online with encryption Obviously so there are different versions you can

actually back up a key and once that is done honestly nobody can touch your files

other than yourself and just don't lose that

and if you lose that there's nothing we can help you

yeah for sure exciting well wishing you and the team all the best in pushing AI agent trading forward I'm going to check the whale list and see who I know cuz it's a

V points plus USDC is much more generous than a lot of these airdrop campaigns have done in the past. You really don't know what you're getting. So yeah, I'll definitely check that out and leave it in the description box below. And the links to the app on the app store also. Wishing you and the team all the best moving forward. Have a strong Q4

ahead and would love to follow up in the near future before the token launch for sure.

Okay. Thank you Ashton for having me. Appreciate everyone. Bye for now.

Related coverage

More interviews

Browse all 1,088 interviews

Get new interviews firstCCS Insider, the free newsletter from Ashton Addison. Twice a week.