For payment processors · merchant onboarding + monitoring
Merchant sanctions screening for payment processors
Payment processors carry hard liability: process a transaction for a sanctioned counterparty + the regulator's call comes the next morning. The screening cost has to be sized for both onboarding (one-off, manageable) and per-transaction (high-volume, demanding). Per-call pricing at the sanctions vendor's retail rate is unaffordable at million-transactions-per-day scale; tools that gate at 5-10s of API latency don't fit payment flows. Stratum runs sub-100ms p95 on a flat monthly subscription that scales to 100,000 checks at £399/mo, with a custom enterprise tier above that.
Scale economics
The published tiers run from Starter (£39/mo, 1,000 checks) to Pro (£399/mo, 100,000 checks, effective ~0.4p/check). A processor onboarding 500 merchants/day + screening 100k transactions/day (~37M screens/year) is well past the published tiers; that's a custom enterprise tier sized to your volume, still a fraction of the £200-400k/year SmartSearch enterprise runs for similar coverage, and without our latency profile.
Daily-list-refresh + audit-trail
The OFSI / UN / EU consolidated lists refresh daily via a cron job; the OpenSanctions PEP catalogue refreshes weekly. List-version timestamp in every response confirms which version your screen ran against; that's the FATF audit-trail requirement. For monitoring (re-screen on existing customers as the lists update): the watchlist tier handles up to 5000 entities at the £49/mo tier; talk to us about a custom enterprise tier.
Dedicated capacity option
For the largest processors (>10k checks/day), Pro tier offers dedicated tenancy + warmup so cold-start latency is never on the critical path. Custom SLA: 99.99% with a sub-50ms p95 ceiling. Named-contact support for incident response. On-prem deployment is on the roadmap for processors with strict data-residency rules + the resourcing to host a Stratum-built Lambda.
Frequently asked
What latency should I budget for in a payment flow?
Sub-100ms p95 on warm matchers + ~30ms for the API gateway = sub-130ms typical. For a payment flow that already runs 200-500ms end-to-end, screening is invisible. Pro tier dedicated capacity: sub-50ms p95.
Do you support transliteration variants?
Yes. The matcher handles Cyrillic-Roman, German umlauts, Arabic transliteration. Threshold defaults to 0.7; pass `?threshold=0.85` for stricter matching at higher precision in payment flows.
How do I monitor existing merchants?
The watchlist endpoint (POST /v1/sanctions/monitored) adds a merchant; daily-cron re-screens + alerts on new matches via webhook (talk to us about webhook delivery for >100 webhooks/day).
List-refresh cadence?
OFSI / UN / EU: daily at 04:00 UTC. OpenSanctions PEPs: weekly Sunday 03:00 UTC. Adverse media: weekly with the PEP refresh. Each response carries the list-version timestamp so the audit trail is honest about which version your screen ran against.
On-prem option?
Roadmap. We're a small team; on-prem is the resourcing question rather than the technical question. Talk to us about Pro + we'll quote the custom build.