Engineered to feel like infrastructure.
The platform review should be short. Two endpoints, one stream of events, and a boundary that stays legible from the first quote to the last payout.
POST /quote → locked, bounded terms (< 200 ms)
POST /transaction → acceptance + replay guards
GET /pool/:partner → your segregated exposure
GET /settlement/:id → audit trail + payout state
webhook: settlement.* → streaming lifecycle eventsCalculation agent, by design.
We don't hold customer funds. We don't route payments. We don't sit inside your regulatory perimeter. The line between your platform and ours is the product.
- Your custody story doesn't change because we shipped.
- Your treasury team keeps signing authority on every hedge.
- Your compliance wrapper applies unchanged. We're software.
- We never touch float. Premiums move through your rails.
- Quote evidence is persisted before any external write.
- Replay protection is structural, not a runtime afterthought.
Small public surface. Deliberate private boundary.
A platform team should leave this page knowing where Slice fits, not how the private machinery works. Exact routes live in docs; venue execution and pricing mechanics do not.
Quote, accept, read back.
The wallet backend prices a specific exposure, presents one customer decision, then confirms the accepted protection after its own premium collection step.
No operational playbook in public.
Public pages can show the boundary. They should not reveal pricing models, exchange routing, hedge timing, or partner credential mechanics.
One partner, one pool.
Exposure is aggregated within your partner boundary. Never across. That line exists in the API, in the database schema, and in every piece of reporting we surface.
Pools are keyed on partner, asset, and terms. Nothing crosses. Nothing commingles. Your exposure is your exposure.
Partner and operator reporting routes are split. Bearer-token auth enforces caller context before any evidence, settlement, or payout view is read.
Review the surface. Bring a flow.
The public docs are enough to evaluate fit. The pilot conversation is where we translate that into a concrete integration plan.