Open source · Calibrated for Kenya VASP Act · Nigeria ARIP · Celo / MiniPay
Pre-launch · In build · Stablecoin Compliance · Africa

Local patterns first.
Global coverage second.

Generic compliance tools learned on Western exchange data. The flows on the ground in Lagos, Nairobi, and Accra look nothing like that. We're building behavioral intelligence calibrated for emerging market stablecoin flows. The thesis and architecture are done; the live API isn't up yet. Before we ship, we want to talk to the operators who'd use it.

In build, not live yet. Open source when it ships. Not a black box.

Opening conversations with operators across Kenya, Nigeria, and Ghana.

Western ground truth is wrong for this market.

Chainalysis, TRM, and Elliptic built their models on Western exchange data. The flows they learned from look nothing like the flows on the ground in Lagos, Nairobi, or Accra. Legitimate flows get flagged, real fraud slips through, and your compliance team burns hours on false positives. Below are two patterns we think generic models systematically miscall — if you operate in this market, we'd like to know whether they match your reality, or whether the bigger problem is somewhere else.

Generic model sees

"This looks like laundering."

A Lagos P2P off-ramp. High-velocity stablecoin movement through informal networks, layered conversions, fast turnover.

What we think it actually is

"This is how dollars cross borders in Nigeria."

P2P off-ramp activity is core stablecoin behavior here. The flow needs to be recognized, not penalized. Tell us if this matches what you see.

Generic model sees

"This is a Sybil ring."

Coordinated deposits from multiple wallets, related on-chain timing, similar amounts. The graph topology screams fraud.

What we think it actually is

"This is a Nairobi chama."

Rotating savings circles are a centuries-old social institution. Coordinated, yes. Sybil, no. We think local social structures should sit in the model as priors — but you'd know better than we would.

What we don't know yet: which of these patterns matters most to your team, what we've named wrong, what we've left out, and whether the cost-of-false-positives in your stack is even the problem worth solving first. That's why we want to talk before we ship — not after.

Forced-buy events the
generic tools don't price for.

If you're a small VASP, on-ramp, or stablecoin operator in these jurisdictions, you have a deadline and a compliance obligation. The enterprise tools don't profitably engage your tier. That's the gap we're building into — and we want to talk to the operators in it before we ship.

Kenya

VASP Act

Kenya's Virtual Asset Service Providers Act establishes a licensing regime for crypto operators with explicit AML and behavioral-monitoring expectations. Existing operators must transition to the new framework.

Transition deadline · 04 Nov 2026
Nigeria

ARIP / ISA 2025

The SEC's Accelerated Regulatory Incubation Program and Investments and Securities Act 2025 set the compliance floor for provisional-license operators. Behavioral monitoring is expected, not optional.

Provisional license framework · active

Pre-launch. We want to talk to you.

Discovery calls are 30 minutes. The product is in build — thesis, architecture, formal specs are done; the live API isn't up yet. We can walk you through what we're building, and we want to find out what compliance teams in Kenya, Nigeria, and Ghana actually need: what generic tools miss, what's working today, and what would have to be true for you to switch.

If your compliance stack is something you've stitched together in-house, or you're staring down the Kenya VASP transition with a small team, or you've looked at Chainalysis and quietly closed the tab when you saw the price — that's the conversation we want.

  • 30-minute call — we have a thesis and an architecture, not a sales deck
  • We ask about your current compliance reality: what you use, what it misses, what your team spends time on
  • You ask anything you want — what we're building, how it works, where we're headed, what we're betting on
  • We share what we're hearing from other operators in your tier — anonymized
  • If a paid pilot makes sense when we light it up, we'll come back to you. If it doesn't, we both move on honestly.

If we're guessing wrong,
we'd rather find out from you.

Behavioral intelligence built from the other direction. In build, not live yet — we won't know if we're solving the right problem until we talk to operators who actually live it.