Omniflow
Omniflow is a payout engine with the documents built in. Paying many people creates paperwork, and today it is put together by hand, after the money is already gone. We make it come out of the payment itself. You upload a list and sign once. Each payment produces its own record: who asked for it, who approved it, the amount, the date and the transaction hash. Before you pay, you can ask a recipient for a W-9 or a W-8BEN, and they fill the official IRS blank. At the end of the year the 1099-NEC comes out filled from the payments themselves.
Videos
Tech Stack
Description
WHO IT IS FOR
Finance teams at companies that pay many people in stablecoins. Today they pick who gets paid in a spreadsheet, chase a tax form from each person by hand, send the money, and then match it all up by transaction hash at the end of the quarter.
WHY NOW
Two things changed this year:
Coinshift, a payout service that had moved about $1.5B, closed its payout product, and its customers were left without a tool.
And since the end of last year, when a company in the United States pays a contractor, that payment is reported on its own, not at the end of the quarter. More payments, smaller payments, and the paper for each one still made by hand.
WHAT PEOPLE USE TODAY
A free Safe sends money to addresses and leaves no paper at all.
Dots, Tipalti and Rise collect a W-9 or a W-8BEN and file the 1099, but they hold your money, they are funded in fiat, and they ask for the form once, when the person signs up.
Request Finance is accounts payable for crypto, with approvals. Its own help centre says an approval rule supports two parameters, Amount and Tag. Searching it for W-8BEN or 1099 returns nothing.
Den and Fordefi let you pay from your own wallet, and they gate on amount, asset, destination and signers. Never on a document.
So today you can have the paperwork, or you can keep your own wallet. Not both.
WHAT WE BUILT
One list. One signature. The money goes out, and each payment produces its own record, as a page and as a PDF.
The documents are part of the same flow. The payer requests a W-9, W-8BEN or W-8BEN-E from a recipient. The recipient picks the one that applies, downloads the official IRS blank unchanged, and uploads the signed PDF. At the end of the year, Form 1099-NEC Copy B comes out filled for every recipient who has a W-9 on file, with box 1 taken from the payments themselves.
We never hold the money and we never sign for the customer. The customer signs, and we propose the batch. That is a product decision, not a legal footnote: since March 2026 the European regulator treats moving tokens on behalf of a client as a payment service, and we stay outside that line on purpose.
We do not file anything, we do not withhold, and we do not tell anyone which form applies to them. Copy A of the 1099 is never produced, because the IRS does not accept it printed from a download.
BUSINESS MODEL
A subscription paid by the payer. Base $100 a month, Pro $500, custom above that. The person being paid never pays.
We take no cut of the money that moves. Moving money is free next to a Safe, and nobody will pay for it.
WHY ARBITRUM
One payout is many transfers, so cheap and predictable gas is the difference between a batch worth running and one that is not. And the receiver leg we are building toward puts a recipient's money straight into whatever protocol they chose, which needs those protocols in one place.
WHERE IT STANDS
This is a pre-production build, put together from working parts to be shown rather than sold.
We are not satisfied with it, and we are already rebuilding. The next version is document first, not crypto first: a new frontend with an interface a finance person can read, and the paperwork, rather than the wallet, as the thing the product is organised around.
The local expertise behind those documents comes from a partner who already does that work. We are a technical team, not an accounting one.
PROGRESS DURING THE BUILDATHON
These three weeks went into understanding the market, not only into shipping features. That was a choice, and it changed the product twice.
Research and customer development.
We built a table of 649 people and organisations who described their own money problem in their own words, each row with a quote, a date and a link.
We wrote 17 messages by hand, got 3 substantive replies and had 4 conversations with people who move money for a living.
One of those replies killed our first idea: the founder of a payout product that had moved about $1.5B said in public that they deliberately never charged for payouts, because a free Safe sits right next to them. That is why we do not sell the transfer.
What the feedback changed.
Session 1: we were told the market we had picked was small and hard, and to find a different ideal customer, so we scored our directions against each other and came back with five open hypotheses instead of one claim.
Session 2: three questions, about regulation, about the tax trail of moving stablecoins, and about how we fit a treasury stack a company already runs, and the record became per payment.
Session 4: we were told to name one region, then decide whether to build the off chain part ourselves or plug into someone who already does it.
What we built.
A contract with no admin withdrawal path and six external state changing functions, a backend, a payer console and a claim page, written from scratch with nothing carried over from our earlier project.
Then the document layer on top: a payment record as a page and as a PDF built with pdf-lib, tax form requests using the official IRS blanks kept unchanged with their source URL, revision and checksum, and a year-end 1099-NEC Copy B filled from the payments themselves.
Plus a landing page, a live example you can open without signing up, and a demo organisation with three months of history.