01 / MONEY

AI Fintech

Payments, wallets and compliance for a world where AI agents pay.

We build the control layer between deterministic financial rails and non-deterministic agents — intent, limits, approvals, receipts — so an agent can spend and a human stays in charge.

Build on our rails ↗
THE CONTROL LAYER
  1. 01Intent

    What should happen

  2. 02Policy

    Limits & permissions

  3. 03Approval

    A human stays in charge

  4. 04Execution

    Deterministic rails

  5. 05Receipt

    Evidence of the result

This is the important bit.
WHY WE DO THIS

This is where the Lab has the most built and the most to prove: working code, first users and a fintech ecosystem around us with real rails and licensed partners. It leads to one position — the trust and delegation layer for agents’ money — and to products that can become companies of their own.

THE WORK

Products in this direction.

01 / AI FintechPilot

MagicPay

Agent payments with a human in charge: a wallet, funding, spend limits, approvals and receipts for AI agents.

In private validation with first users.

Ask about MagicPay ↗
02 / AI FintechPrototypetestnet

X402 Hub

An open protocol layer for agent-to-merchant payments: buyer vault, MCP and CLI adapter, EVM contracts, Stellar testnet, merchant registry and bond.

Reproducible end to end on testnet.

Ask about X402 Hub ↗