Cold Storage, Designed for the People Who Actually Use It

Most cold wallets ask too much of the people they are trying to protect. A 24-word phrase on paper, locked in a drawer, copied wrong, lost. The friction is the failure point. By the time the wallet is needed, the recovery path is already broken — words transposed, the page water-damaged, the safe combination forgotten.

The Tychi Cold Wallet is a response to that. We did not invent better cryptography. We changed how the recovery path is split between hardware and memory.

The split model

Tychi uses a 15-word setup, intentionally divided:

• 12 standard seed words stay on the NFC card. Never on the phone.
• 3 personal words, chosen by the user, kept in their head.

Access requires both. The phone alone cannot unlock the wallet. The card alone cannot move funds without the user's three words. To anyone who steals only one half, the other half is missing.

What this is, honestly

We want to be careful here, because the framing matters.

This is not a cryptographic improvement over a 24-word seed. The standard 24-word phrase already provides 256 bits of entropy — far more than three additional user-chosen words can add. Three personal words from a small mental dictionary contribute very little to brute-force resistance. We are not claiming the math is stronger.

What it is: a usability split. The 12 dictionary words live on hardware that is hard to lose, hard to read without a phone, and hard to copy by glance. The 3 personal words live in the user's head — somewhere a phone hack cannot reach. The combination shifts the failure mode from "user mishandles a paper backup" to "attacker needs both the physical card and personal information they don't have."

For some users, that trade is the right one. For others, a screen-equipped device with full transaction confirmation will be the right answer. We are not arguing one model fits everyone.

Why split makes sense for our user

The user we are designing for is not the hardware-wallet hobbyist. It is the person who buys a hardware wallet, sets it up, and never touches it again until they need to move funds — at which point the recovery path has degraded. NFC + 3-word UX is built for that user.

• If a phone is stolen, the card tap is still missing.
• If a phone is hacked, the attacker still does not have the user's three words.
• The split adds a physical and a memory layer on top of what the user already knows.

The hardware path

The product is shipping soon. When it does, two paths will work:

• The Tychi cold wallet card, ordered directly from us. Branded, engraved, ready to pair.
• A fresh NTAG215 NFC card, the same chip used in many third-party cards. No vendor lock-in. If we go away, the cards still work.

The choice is intentional. Closed hardware platforms are a form of risk. If a vendor disappears, paired cards become bricks. By supporting the open NTAG215 standard, the Tychi flow continues to work even without us in the loop.

What ships with it

• Pair-and-go setup in the Tychi Wallet app.
• Tap-to-sign for any transaction the wallet authorizes.
• UGF integration on the same flow — pay gas in stablecoins from a cold-signed transaction.
• Recovery via the card plus the user's 3 personal words.

What does not ship with it

We are not putting a screen on the card. We are not building a custodial recovery service. We are not selling a multi-thousand-dollar institutional variant. The product is a sub-$100 NFC card paired with the wallet.

If you want to be notified when it ships, head to tychiwallet.com/cold-wallet. The waitlist is real, and the people on it get the first cards.

The 12+3 model is the most honest answer we have to a real problem: most users do not back up their seed phrase correctly. Splitting the burden between hardware and memory will not fix that for everyone, but it shifts the failure mode toward something more recoverable. That is the bet.