The autonomous economy already exists. The bank it needs doesn't.
Vayan exists to close the gap between what agents can already decide and what today's banking infrastructure still makes humans approve manually.
Why this exists
The click that breaks autonomy
An AI agent can negotiate, procure, route, and decide. Then it reaches the payment step and stops because a human still has to open a laptop and click approve. The autonomous economy already operates. The bank it needs does not exist yet.
The bank-side gap
Regional banks have trust, clients, and regulatory standing, but not the architecture for continuous examination, dual-rail settlement, or agent-operated commerce. They have the asset. They lack the operating layer.
The infrastructure-side gap
Software-native workflows increasingly originate financial instructions programmatically. The hard part is not adding an API. It is keeping governance and auditability intact while the work moves to agents.
We noticed the same thing from different sides.
Hubert spent years inside regulated banking operations. He saw how compliance, settlement, and reporting still depend on people manually coordinating processes that could be structured, governed, and performed by software if the architecture were designed for it from the start.
Yves came from institutional markets and infrastructure. He saw how financial activity is becoming increasingly software-originated — agents making decisions, protocols settling value — and how existing banks cannot absorb that shift without changing the operating model underneath.
The overlap was the founding insight: the bank of the future does not need better tools for humans. It needs an operating layer that performs the repetitive work directly, under structured governance, with continuous evidence.
What we learned before building
Every design decision in Vayan came from something that failed or fell short in prior work. These are not theoretical preferences. They are conclusions from operating experience.
You cannot retrofit compliance into speed
Every bank that bolts compliance onto a fast settlement layer ends up with a compliance bottleneck. The only architecture that works is one where policy validation is part of the execution path itself.
Software vendors create dependency, not transformation
Licensing copilot tools makes the vendor indispensable without changing the bank's operating ratio. The model that works is shared operating savings: Vayan only makes money when the bank's cost structure actually improves.
Regional banks are the right starting point
Tier-1 banks have the budget to build internally. Neobanks start from scratch without a licence. Regional banks — 4,500+ in the US alone — have the licence, clients, and capital but not the platform. The network model is designed for them.
Regulation is a moat, not a burden
Every competitor that treats regulatory authorisation as a cost to minimise is building on sand. Vayan treats it as the primary design constraint because the institutions that matter will only work with counterparties that take it seriously.
The decisions that shaped the architecture
Perform the work, don't sell tools for it
Revenue comes from operating savings and transaction throughput, not licence fees. If the bank's cost structure doesn't improve, Vayan doesn't earn.
Build compliance into execution, not around it
Policy validation happens before every regulated action. Evidence is generated as a byproduct of execution, not reconstructed for auditors after the fact.
Dual-rail from day one
Fiat and stablecoins are two rails of one balance sheet. Building one and adding the other later produces architectural debt that never gets repaid.
Human governance is structural, not optional
Multi-Agent Consensus means no single agent can execute alone. Material transactions route to named human officers. The policy layer cannot be bypassed.
The network compounds structurally
Every additional bank adds transaction data, which improves compliance intelligence, which lowers cost per bank, which attracts more banks. The economics improve with scale by design.
The founding team
Hubert Knapp
Previously built and operated regulated banking infrastructure. Approaches Vayan as a prudential and regulatory design problem first, not a technology problem. Responsible for the control architecture, partner-bank operating model, and the governance framework that ensures human authority over agent execution.
Yves R. Burri
Background in institutional markets, capital strategy, and infrastructure architecture. Leads the thesis that future financial activity will be increasingly software-originated and that banks need an operating layer designed to absorb that shift under continuous regulatory scrutiny. Responsible for the capital strategy, institutional positioning, and agent-banking interface design.
What we are building toward
arrow_forward A dual-rail operating layer where fiat and stablecoins behave as one balance-sheet reality.
arrow_forward A network model in which partner banks keep the licence and Vayan performs the repetitive work.
arrow_forward A control model where policy validation and structured evidence sit inside execution rather than outside it.
arrow_forward A data moat where every transaction processed makes the network more intelligent and more defensible.
The window is open because the architecture changed.
Stablecoins gained a clearer regulatory path. Language models crossed the threshold for multi-step financial reasoning. Institutions now need a bank that can operate on both rails with continuous evidence. This is the moment to build it.