A shop proves it is real with a zero-knowledge proof, takes an advance in stablecoins, and repays 8% of every sale until it is cleared.
No collateral. No credit file. No audited books. A kiosco in Buenos Aires, a stall in Lagos, a duka in Nairobi.
Divine reported roughly 40% defaults on first loans, even with World ID blocking repeat borrowing.
Shopify Capital advanced $4.2B this way in 2025. Our version needs no employer and no partnership: the checkout link collects.
A zero-knowledge proof of the seller dashboard: country, revenue band, six months of trading. No name, no phone.
Customers pay in stablecoins. No checkout fee. While an advance is live, 8% of each sale goes to repayment.
Limits set once, enforced onchain. An agent draws small advances to pay suppliers without asking again.
A cleared advance mints a private credential that raises the limit. No history is revealed.
Most companies at this stage raise to build. We raise to find out whether merchants repay.
| Time outstanding | Fee on $100 | Total repaid |
|---|---|---|
| 1 week | $0.50 | $100.50 |
| 4 weeks | $2.00 | $102.00 |
| 6 weeks and beyond (cap) | $3.00 | $103.00 |
With no discount for paying early. Ours stops accruing the day principal clears.
Capped at 6%. All of it funds the buffer that absorbs defaults. On-time merchants never see it.
Losses hit the junior tranche first, then the reserve, then senior. No validated loss history yet.
Nobody changes how they get paid for a better rail. They change it for working capital they cannot get anywhere else. The advance is the reason to switch.
One city with stablecoin habits and a licensed off-ramp. Ship live seller data. Hand-onboard 25 shops against a small pool.
We measure four things nobody has for this model: loss rate, repayment pace, time to first advance, dropout per step.
Wholesalers want paid-on-delivery orders. A supplier paid upfront brings us the shop. Second advances are bigger and faster.
Open the senior pool to outside lenders once the loss series is published.
New seller data, new off-ramp, nothing new onchain. Audit and key hygiene before meaningful outside capital.
| Layer | Number | Source |
|---|---|---|
| The need | $5.7T finance gap, 70% of small businesses underfinanced | IFC |
| The model | $20B to $46B advanced a year, mostly US. Shopify Capital $4.2B in 2025 | Mordor, Shopify |
| Incumbents in our markets | Moniepoint about $700M to small businesses in 2025 | Launch Base Africa |
| Bottom-up | 100k merchants, $200 average, 6 cycles a year: $120M advanced a year | our assumption |
Onchain credit scoring is crowded. Merchant advances in these markets are empty. We never pitch this as a credit score. It is an advance against a flow we can already see.
Contracts, proofs, agents and app. No agency, no contractors, no outside engineers. Built at ETHGlobal Mumbai 2026 and open-sourced the same week.
We are not raising to build the product. It is built, deployed and open. We are raising to find out whether real merchants repay.
| Use of funds · 18 months | USD |
|---|---|
| Team: four people, 18 months | 190,000 |
| External security audit | 60,000 |
| Pilot market ops and onboarding | 60,000 |
| Live seller data, off-ramp, infra | 40,000 |
| Entity, legal, compliance ×2 markets | 30,000 |
| Buffer | 20,000 |
| Total | 400,000 |
Around a $5M post-money cap. A $90K to $500K first cheque fits as lead or co-lead.
Public-goods work that stands on its own, grant-shaped rather than equity-shaped.
First-loss capital so 25 merchants can draw real advances. From lenders or a grant, never out of the round.
A public ledger of which shop borrowed how much is a safety problem in these markets, not a transparency win.
Every transaction is public. Proof verifications and AI decisions are readable by anyone.
Seller identity and revenue detail are proven, never sent. Passes are proven in the browser. Private reads by default. No trackers.
If our services vanish, the contracts carry on. Non-upgradeable. No pause.
The whole stack builds on a laptop with no real keys.
Device-side proving and private reads on by default, not a setting to find.
Caps and mandate limits onchain. Agents only propose. This is unaudited pilot code.
The junior tranche trades against stablecoins on Uniswap v4. If the market marks it down, merchant limits come down before losses land.
Three quarters of it funds first loss, a quarter is treasury.
The risk agent can only lower, never raise, and never below the floors.
The pool is thin and team-seeded, which is why write-offs and repayment pace are the other two triggers.
Not as a trade. As money. And the markets where credit is scarcest are adopting fastest.
A mid-range Android can generate them, and an existing seller dashboard becomes a credential with no API deal.
A merchant delegates once under onchain limits, so an agent can pay suppliers without a tap each time.