Platform

Three systems. One continuous read on every identity you hold.

Onboarding Shield stops synthetic identities at the door. Sleeper Graph Engine watches the ones that get through. Case Intelligence gives your investigators a defensible case in minutes, not weeks.

01 — Onboarding Shield

Catch fabricated identities before they open an account.

At the point of application, Onboarding Shield runs document forensics, active liveness and deepfake-injection detection, and cross-bureau plausibility scoring on every identity element — SSN issuance pattern, address history depth, name-to-age consistency. Applications score in under 400ms, so the check never adds friction to a real customer's application.

Document forensics Liveness & deepfake detection SSN issuance-pattern check Bureau cross-plausibility
Sample decision
APPLICATION_ID: 7731-A2
SSN_ISSUANCE_MATCH: PASS
DOC_FORENSICS: PASS
LIVENESS_SCORE: 0.97
ADDRESS_HISTORY_DEPTH: LOW — 41 days
NAME/AGE PLAUSIBILITY: FLAGGED
————————————
DECISION: MANUAL REVIEW
LATENCY: 312ms
Sample graph read

Identity 84041-C shares a device fingerprint with 11 other accounts opened across 4 institutions in the last 90 days. 9 of those 11 show the same address-seasoning pattern: 3 address changes in 18 months, each held just long enough to clear a bureau refresh cycle.

CLUSTER RISK: 0.91 · RING SIZE (EST): 30–45 IDENTITIES

02 — Sleeper Graph Engine

Find the ring, not just the identity.

Synthetic identities are rarely built one at a time — they're manufactured in batches from the same playbook, sharing devices, addresses, and application timing patterns that are invisible within any single account. Sleeper Graph Engine continuously links identities across your full portfolio and, where you participate, across the Wakeful consortium network, surfacing clusters that share a common origin before any single member of the cluster activates.

03 — Case Intelligence

Every flag arrives explainable, evidenced, and exam-ready.

A flag your investigators can't explain to an examiner isn't actionable. Case Intelligence packages every score with the underlying evidence graph, a plain-language rationale, and a model documentation trail aligned to your SR 11-7 / OCC model risk management obligations — so a flag becomes a decision in minutes, not a two-week investigation.

What ships with every case

Linked identity graphIncluded
Feature-level rationaleIncluded
Model documentation trailIncluded
Suggested SAR narrative draftIncluded
Analyst override & feedback loopIncluded

Most teams run a 60-day shadow deployment against their existing book before going live.

Ask about a shadow deployment