Identity is a pattern, not a moment. Prove you're human in twelve seconds.

On-device behavioral liveness, anchored on Robinhood Chain. Proof-of-personhood for humans, accountable identity for the agents acting on their behalf. No raw biometrics.

Coming soon
1
The problem

You can no longer tell a human
from a bot on-chain, and finance is where that costs the most.

Airdrops get farmed by a thousand scripts. Governance gets captured by whoever spins up the most wallets. And a growing share of on-chain capital is moved by autonomous agents with no accountable human behind them.

Balances can't answer the live question, balances are for sale. KYC can't either: it proves an identity existed at signup, not that a human is present now, and it hands a third party a pile of PII in exchange.

Voice, motion, and touch captured on-device for twelve seconds

A zero-knowledge proof matches the fingerprint to your wallet's prior sessions

A non-transferable credential is anchored on Robinhood Chain

2
Protocol

Temporal identity, built on Robinhood Chain.

Behavioral capture

Voice, motion, and touch captured together against a fresh random challenge each session. Raw data never leaves the device.

Groth16 / BN254

A zero-knowledge proof that the current fingerprint sits within human-drift range of the wallet's stored commitment.

Soulbound Anchor

ERC-721 with ERC-5192 lock. Personhood is bound to the wallet and cannot be bought, sold, or transferred.

Liveness model

Behavioral-entropy scoring and synthetic-data detection decide whether a take is a live human or a generated one. This is the moat.

Operator Bond

An autonomous agent bonded to a verified human's Anchor. The agent acts freely; accountability traces back to a real person.

Score-weighted governance

An EVM Governor with a voting-weight strategy that reads the Tempo Score instead of token balance. One human, one vote.

4
How it works

Capture. Prove. Anchor.

STEP 01

The Take

Voice, motion, and touch captured together against a fresh random challenge each session. Raw data never leaves the device or persists, only derived statistical features are checked, and entropy scoring flags synthetic patterns on the way in.

STEP 02

The Proof

A Groth16 proof that the current behavioral fingerprint sits within human-drift range of the wallet's stored commitment. A minimum-distance constraint kills perfect-replay attacks. No raw biometric is ever exposed or written on-chain.

STEP 03

The Anchor

A non-transferable credential bound to the wallet, settled on Robinhood Chain. Verify once to register; verify again to prove you're still you.

STEP 04

The Score

The Tempo Score rises with consistent re-verification and decays with dormancy, so it reflects sustained real use rather than a single event. A fresh sybil has no history to imitate.

Testnet · Robinhood Chain

Prove there's a returning human here.

Twelve seconds of voice, motion, and touch, captured on your device, proven in zero knowledge, anchored to your wallet. No raw biometrics persisted, no PII collected.