The Modular Hardware Wallet
- Guy-Louis Grau Product Lead · Keycard
Crypto has always had a sense of freedom about it. The freedom to transact whenever you want, on whatever chain you want, however you want. Keycard is taking this freedom to the next level - bringing modular hardware wallets to ordinary users.
Transcript
Hodl Up: Hello guys and welcome back to another episode of Hodl Up. Today we have GL from Keycard with us. Thank you for joining us today, GL. Give us a TL;DR of what you're building right now.
Guy-Louis: I'm GL, and I'm the lead of a project called Keycard. We make hardware wallets, and what matters to us is bringing openness and modularity to how hardware wallets are made. We build technology, and we also build products. Our story is that we think hardware wallets are currently very monolithic. They're very often closed source, so you don't know what's happening in your hardware wallet, and they often pull you into their own ecosystem. You need to use their application, and in that application you're very often being sold things, swap services and so on. What we built is a system based on smart cards, called Keycard. This card is very secure. It has a certified secure element that holds your keys, and your keys never leave the card. Once you've set up your card, no one will ever be able to get the keys out: not you, and not us as the manufacturer, even if we decided to or were forced to. So it's all about giving control to users. Once their keys are here, they decide how to use them.
With this card, you can of course use it directly with a mobile phone. More than 15 wallets on mobile and desktop work with our system, so you can tap the card to sign a transaction. That's one way to use it, on the go. It's NFC: no battery, it always works, no pairing. It's a great experience. You can also use it on desktop with a desktop wallet, for instance Sparrow for Bitcoin or Status Desktop for Ethereum. You use any USB reader that works with smart cards, plug it in, and when you want to send a transaction, your wallet communicates with the card.
Finally, and probably most interesting, we've built a hardware wallet called Shell, which is a way to bring interfaces to this card. When you put a card in, you get a full secure interface for it. For instance, I enter my card's PIN here, and then I have a full-fledged hardware wallet with a camera. It's an air-gapped hardware wallet, with no Bluetooth or Wi-Fi communication, so it's very secure, and you can use it with a lot of different wallets, like MetaMask, Rabby and imToken. There are about 15 of them, because it uses QR codes. There are standards for carrying transactions over QR codes, and a lot of wallets have adopted them. So let's say I want to use the keys on my card with the Shell and a wallet. I'll use my mobile here, but I could use my desktop. I open a wallet like MetaMask, just to show you the principle. I get my regular experience for creating a transaction. You won't see it here, but I pick some assets to send and create the transaction, and MetaMask shows me a QR code. I grab my device, press one button, scan the QR code, and I get the full transaction on this screen, which is a secure, air-gapped screen. I can check all the details of my transaction. We do clear signing, which shows all the details of what you're signing. When I'm happy with the transaction, I sign it with the card, and this is my signature. I send it back to MetaMask by scanning it with the app. That's the whole system, where owners truly control their keys, because they decide how to use them. The keys aren't locked into your hardware wallet. We're bringing more and more wallets into the ecosystem, and we want to diversify and bring new use cases. These cards are also very well suited to payments, for instance, so that's something we're working on.
Hodl Up: First of all, thank you for the whole demonstration. You're the first builders who have actually shown us the whole journey and process of how everything works, and it looks super cool. What led you to create this product? What was the decision point where you said, "Okay, people need something like this"? Personally, I've been scammed and lost quite a lot of money in the past. I've had wallets drained, for example. But I'm not a builder, so I could never have built something like this. I want to hear from you.
Guy-Louis: The starting point is really self-custody. We started this project about seven years ago, in 2019, and it has always been very important to us to explain why self-custody matters, because a lot of users come from worlds where their passwords are centralized and they rely on someone else to manage them. So we wanted to do something about self-custody. That's the first thing. And self-custody needs hardware. When your private keys live in software on your phone or your desktop, you're at risk, because those are insecure environments. So we wanted to do hardware self-custody, and the main thing was to do it open source. We really took open source as the main constraint. When you're making hardware wallets, you first need a secure element, a very safe place to store a secret. So our first decision was: what's the most open secure element on the market? That led us to Java Cards. Java Card is a technology everyone on the street carries with them, because every bank card and every SIM card in your phone uses it. It's very mature and very secure, and the beauty of it is that it's an open platform where you can run open-source applets. So, to answer your question, the first part of the journey was choosing smart cards, Java Cards to be precise, as secure elements. They're great, but they lack interfaces. You don't have a screen, and you want to check what you're signing. So the second step of the journey was to say, hey, we need to bring a screen and a keypad to these cards. That's why we built this, and we wanted it to be open source and modular, which explains all the choices in how it was built. For instance, the battery is removable, so it won't die on you after three or four years. If the battery gets too old, you can replace it. It's a standard battery; we use the battery from Nokia phones, which are sold worldwide, so you can buy a replacement in any country, swap it, and your product works again. The whole idea is to create something where users decide. One of the unique things we do is that with this type of card you can actually build your own card. Since these cards are everywhere, you can buy a blank white card without any software on Amazon or AliExpress, take our software from GitHub, verify it if you're a developer, and load it onto the card. Let's say you have millions of dollars to manage and you want to be sure of the exact software running on your secure element: here you can go through every step and put the software on the most important part of the system yourself. That's why everything is open source. It's not security theater where just one part of the software is open source. You can find all the software, for the secure element and for this device, plus the whole hardware design, the schematics and layouts. If someone wants to build their own device, it's really possible, and some members of our community are building their own plastic cases. So that's why we brought this to life.
Hodl Up: I think you should be hired in sales as well, because you just sold this product to me. It even looks like one of the safest products out there, or maybe you just sold it to me, as I said. I'd like to know about your users' demographics in terms of profession. Are they more tech-proficient, or are they just random people who want to keep their money safe?
Guy-Louis: First, it's a little hard to know, because we don't track anything about our users. When you have your own wallet, we don't track who's using it or what they hold. But from the questions we get from people who bought the product: we're a new project, we don't have a well-known brand and we're not everywhere, so it turns out it's mostly tech people who understand the value of open-source, modular technology. Tech-driven people, lots of crypto natives, lots of people who already had a hardware wallet and had bad experiences. So most of the people we talk to today are from this industry, developers or tech-aware. We think the natural next step will be reaching less crypto-native people, mostly because of the convenience. There's convenience in having several wallets and several cards, for backup for example. We ship this product with two cards, but you can buy more, and it's very convenient to have several cards as backups: you just insert your card and you're ready to go. So we think it's a very convenient product, and the time will come when we'll have normies.
Hodl Up: Can I pick the design of the card, or is it pre-selected?
Guy-Louis: For now we've made it as stealthy as possible. We ship a black card with the product. There's no name on it and absolutely no branding on this device, because the priority was to make it really stealthy. If someone finds this card in your wallet, they won't immediately guess that it's a backup or a signer.
Hodl Up: That brings me to my next question, because that's really important. A wallet shouldn't scream, "Hey, I'm rich, please come rob me." At the same time, you mentioned the card being super convenient, and I agree. I've used different wallets in the past, and because they're not very convenient and sometimes bulky, I've never left my house with a hardware wallet. I always keep them somewhere at home. But in this case I'd be very inclined to keep it in my wallet, because it's convenient and small. I have a card-style wallet that fits a lot of cards. But that would also make it more likely to be stolen, precisely because it's so convenient to carry around. So how do we let people know that maybe they should keep these things at home? You said the design is intentionally simple and not flashy, but I want to hear your take, as a potential user who might have this concern.
Guy-Louis: There are a few things to mention. First, the card is protected by a PIN code, and it gets blocked completely if you enter the wrong PIN more than three times. That's what protects it against attacks. If someone steals your card, they get three tries and then it's completely blocked. The second thing is a system called duress PIN. You can set up your card with a second PIN that opens another wallet. That's important if you're being attacked, for instance, and someone grabs your key, demands your PIN and wants to be sure you gave them the right one. You'd enter your duress PIN, and it would still open a wallet with some funds, so you could show that you didn't cheat the attacker. That's the second security mechanism. And yes, one of the things we'd like to do is print custom cards that are even more than stealthy: cards that look like gym cards or membership cards. There's no limit to what we can do. We're not in a position to do that right now because we're launching this product, but it would make a lot of sense. And if someone is really motivated, nothing prevents them from doing it, because you can buy white cards and print them yourself. There's a do-it-yourself side to this product, and we allow it.
Hodl Up: A few minutes ago you mentioned you'd try to target normies as well and bring them in to use your product. I think that's what this industry needs right now, new people coming in, because I believe everyone will appreciate what we're building as an industry, and you specifically. You're building something amazing here. Genius, actually. I fell in love with the product, and I'll keep saying it because it's so amazing. But how do you plan to onboard these normies? We know the average person nowadays doesn't care about privacy or safety. People don't care about anything anymore. I want an idea of how you'd approach that.
Hodl Up: And to add to that: I saw a lot of normies onboarded last cycle, and the ones who were interested in securing their funds with cold wallets would first ask, "Should I buy a Ledger?" How do we convince people that there are alternatives that might be even safer than these massive corporations?
Guy-Louis: To bring in normies, I think there's a UX question, which is important, and everyone knows our user experiences in this industry aren't friendly. One thing we're working on is more clear signing of transactions on these screens. There's a big initiative from the Ethereum Foundation called ERC-7730 to display the intent of the transaction you're signing on the screen. We're part of that group and working to bring it to Shell. That's one thing, and making a better user experience overall is the main thing. Price point also matters for reaching the masses, and that's another reason this form factor is great: these cards are everywhere. Billions of them are manufactured every year, so it makes for a very secure hardware wallet at a very good price point, which matters for mass adoption. And your third point, sorry, I forgot what it was.
Hodl Up: How do we convince new people who want to protect their funds to come to you, and not go to bigger brands, the leaders in hardware wallets?
Guy-Louis: It's all about trust. What we're trying to build is a product you can trust because of the product, not because of the company behind it. You can verify the code, and you can verify that the right code is running here. Everything is open source. You can build your own secure element. You choose the app you use it with; you don't have to download a specific app. So I think it will mostly speak to people who are tired of trusting big brands and want devices that are more do-it-yourself and open source. It won't speak to everyone right now, but that's what matters to us, and that's where we bring a solution.
Hodl Up: I don't know if you expected to create two new customers when you came on this episode, but you did. Now I'm curious: how much does it cost, and does it come with all the features, the Shell and the USB reader, or are they sold separately?
Guy-Louis: The Shell comes with two cards in the box. The price is €149, and we're currently selling it for €129 because we launched it two months ago. So it's €129 with two cards, and the reader is about €20 on top. You can buy as many extra cards as you want. So that's the price point.
Hodl Up: That's quite reasonable compared to what some other providers charge, and they don't even include the USB-C cable you may need. Well, thank you so much for joining us today, GL. As I said, you onboarded two new people today, if not more, depending on whether viewers saw the full demonstration. If they did, I'm sure they'll be hooked as well. Thank you so much for joining us. It was a huge pleasure, and we'll be rooting for you.
Guy-Louis: Thank you. Take care.
Key takeaways
- Keycard stores private keys on a Java Card smart card with a certified secure element, and the keys can never be extracted, even by the manufacturer.
- The card works by NFC tap with more than 15 mobile and desktop wallets, through a USB smart card reader on desktop, or inside the Shell device.
- Shell is an air-gapped hardware wallet with a screen, camera and keypad that signs transactions by QR code with wallets like MetaMask, Rabby and imToken.
- Every layer is open source, and users can buy a blank card, load Keycard's software from GitHub and build their own secure element.
- Security features include a PIN that blocks the card after three wrong attempts and a duress PIN that opens a decoy wallet.
- At recording time, Shell with two cards sold for a €129 launch price (€149 list), with a USB reader around €20 extra.
Hardware wallets without the lock-in
In this Hodl Up episode, Guy-Louis Grau, known as GL, walks the hosts through Keycard, the hardware wallet project he leads. Most hardware wallets are closed boxes: proprietary firmware, a required companion app and, often, built-in upsells. Keycard takes the opposite approach.
"What matters to us is bringing openness and modularity to how hardware wallets are made," GL said. "It's all about giving control to users. Once their keys are here, they decide how to use them." The conversation includes a live demo, the reasoning behind Keycard's design, its security features, who uses it today and how it plans to reach mainstream users.
How Keycard works
The card and its secure element
Keycard is a smart card with a certified secure element holding the user's keys. Once set up, "no one will ever be able to get the keys out: not you, and not us as the manufacturer, even if we decided to or were forced to."
The card can be used three ways. Users can tap it against a phone over NFC, with "no battery, it always works, no pairing." They can plug any smart card USB reader into a desktop and use wallets like Sparrow for Bitcoin or Status Desktop for Ethereum. Or they can insert it into Shell.
Shell, the air-gapped companion
Shell gives the card a screen, keypad and camera with no Bluetooth or Wi-Fi. In GL's demo, he built a transaction in MetaMask, scanned its QR code with Shell, reviewed the full details on the secure screen, signed with the card and scanned the signature back into MetaMask. Shell supports clear signing and works with about 15 wallets that follow QR transaction standards, including MetaMask, Rabby and imToken.
The table below compares the ways to use Keycard.
| Setup | How it connects | Example wallets | Best for |
|---|---|---|---|
| Keycard and phone | NFC tap | More than 15 mobile wallets | Signing on the go |
| Keycard and USB reader | Smart card reader on desktop | Sparrow, Status Desktop | Desktop use |
| Keycard and Shell | Air-gapped QR codes | MetaMask, Rabby, imToken | Maximum security with a screen |
Why open source and modular
Choosing Java Card
The project began in 2019 with a focus on self-custody. GL argued that keys stored in software on a phone or computer are at risk, so self-custody needs hardware, and Keycard set open source as its main constraint. That led to Java Card, the same technology inside bank cards and SIM cards. "It's very mature and very secure, and the beauty of it is that it's an open platform where you can run open-source applets."
The modular approach extends to Shell. Its battery is removable and uses a standard Nokia phone battery, so it can be replaced in any country.
Build your own card
Anyone can buy a blank card on Amazon or AliExpress, download Keycard's software from GitHub, verify it and load it. The secure element software, device firmware, schematics and layouts are all public, and community members are building their own cases. "It's not security theater where just one part of the software is open source," GL said.
Security and stealth
The card ships black with no branding, so a thief would not recognize it as a signer. It blocks after three wrong PIN attempts. A duress PIN opens a separate wallet with some funds, which a user can reveal under threat. GL would like to print cards that look like gym or membership cards, and users can already print their own.
Who uses Keycard, and who comes next
Keycard does not track users, but most buyers appear to be developers, crypto natives and people who had bad experiences with other hardware wallets. GL expects mainstream users to follow as UX improves, including clear signing through the Ethereum Foundation's ERC-7730 initiative, and because mass-produced cards keep prices low.
On competing with major brands, his answer was trust. "What we're trying to build is a product you can trust because of the product, not because of the company behind it."
