Why we are building Ones
Written by the founders · August 2026
We spent years running money for companies before we started this. The same three things went wrong every month, at every size of business, and none of them were anyone’s fault. They were the shape of the tools.
The problem we kept hitting
The first is privacy. When a company banks, its balances, its suppliers and its timing end up readable by more people than the company ever agreed to. Data gets pooled to build products, shared with partners, and sold on in aggregate. For a business with a competitor in the same market, who you pay and when you pay them is strategy. It should stay yours.
The second is foreign exchange. A company paying a supplier abroad loses money twice: once on a spread hidden inside the rate, once on a fee at the far end. On a few hundred thousand dollars a month that is a salary. Most finance teams have no way to see the spread, so it is very hard to argue with.
The third is everything that arrives on the statement after the fact. Wire fees, receiving fees, minimum balance charges, account maintenance, a card scheme markup. Each one is small. Together they are a line item nobody budgeted for, discovered a month late.
What we built
Ones is a bank account for people, for businesses, and for the agents that act on their behalf. We built it in layers, and each layer answers one of those three problems directly.
The bottom layer is the ledger. Every account sits in its own ledger rather than in a shared pool. A payment proves itself to the two parties in it and to nobody else. There is no combined dataset for us to analyse, license or lose, and that is a property of how the thing is built rather than a policy we could quietly change later.
Above that is the money layer. One account holds several currencies, stablecoins, tokenized gold and other real-world assets, and you swap between them in one step. Say you are in India and you want a tokenized share: you buy it here, and you can move it to any wallet you control, because it is a token you hold rather than a line in our database. We show you the rate we get and charge you that rate, and the fee is on the screen before you press send.
On top sits the console. Everything the account holds reads as one balance, whatever it is made of. Payments run in batches, exceptions surface as they happen, and approvals go to the person who is meant to sign them. One screen to check in the morning rather than a spreadsheet.
Where the AI earns its place
A finance team spends most of its week on work that is careful and dull. Matching two thousand transactions to bank statements. Reading four hundred lines to find the one that is odd. Chasing an invoice through three systems to work out whether it was paid. This work has to be right, and doing it by hand is the least reliable way to get it right.
So we automated it. Overnight, Ones reconciles your accounts, matches invoices to what settled, and writes you a short summary of what happened. Anything unusual comes to you as one notification with the entry attached. Agents can hold their own accounts with their own limits, do the routine work inside them, and stop at anything that needs a person. A new supplier and a large amount always wait for you.
Every one of those features sees one slice of your data and no more. The sensitive work runs in hardware where even we cannot read into it. Your book trains nothing.
What it costs
There are no hidden fees. No monthly charge, no minimum balance, nothing added to the exchange rate, and no charge that appears on a statement you have not already seen on a screen. If we ever introduce a price, you will hear it from us first.
Who is building it
Four of us. Ayush Singh and Ankush Singh are the co-founders, Meet Jain is head of engineering, and Harsh is a founding engineer. We would rather publish the names now than a paragraph of adjectives later. If you run money for a company and any of this sounds like your month, come and tell us where we have it wrong.