The Question
What if crypto wallets didn't need passwords or seed phrases at all?
Every existing wallet solution forces users to either trust a custodian (and accept the counterparty risk of an FTX-style collapse) or manage their own keys (and accept the UX burden of seed phrases, hardware devices, and the permanent risk of loss). The industry had accepted this as a fundamental tradeoff.
Project Eden was a weekend hackathon to test whether WebAuthN and PassKeys (the same technology that lets you log into websites with your fingerprint) could generate and manage cryptographic keys suitable for blockchain transactions. Not a wrapper around existing keys. Not a custodial abstraction. Actual non-custodial wallets where the passkey is the private key material.
The Prototype
Over a weekend, I built a working proof-of-concept demonstrating:
- Passkey-derived wallet creation — using WebAuthN credentials to generate deterministic key material for Bitcoin and EVM networks
- Schnorr signature transactions — signing Bitcoin transactions using passkey-derived keys without ever exposing a seed phrase or raw private key
- MPC transaction signing — multi-party computation approach where no single device holds complete key material
- Cross-device compatibility — passkeys sync across devices via platform authenticators (iCloud Keychain, Google Password Manager), meaning wallet access follows the user without manual key backup
- Multi-chain support — same passkey generating keys for Bitcoin (Schnorr/Taproot) and Ethereum (secp256k1)
Why It Works
The insight is that WebAuthN's credential creation ceremony already generates a P-256 or Ed25519 keypair bound to a specific origin. With the right derivation path, this can produce keys compatible with blockchain signature schemes. The user never sees a seed phrase because there isn't one. The key lives in the device's secure enclave, protected by biometrics, and synced via the platform's existing passkey infrastructure.
This isn't custodial. No server ever holds the key. It isn't traditional self-custody either. There's no seed phrase to lose. It's a third category: platform-secured self-custody, where the security model is the device manufacturer's secure enclave rather than the user's ability to remember 24 words.
What Happened Next
The prototype was functional enough to demonstrate the viability of the approach. I co-founded Oviato to turn this proof-of-concept into production infrastructure: passkey-native wallet-as-a-service APIs for third-party builders. The technical demonstration and pivot narrative directly secured DraperVC investment.
Project Eden proved that the hardest problem in crypto UX (key management) had a solution hiding in plain sight in the Web Authentication standard. Oviato is packaging that solution for the industry.