Solutions, By function, SIU & Fraud

Every pattern, on every file

Referral rates run one to three percent of files. Actual fraud runs eight to ten. Owl reads every file at FNOL against the patterns your SIU team has chased for years, and the patterns nobody on your team has seen yet, because it was a peer carrier that investigated them first, not you.

Where fraud actually hides, cohort n=2.4M filesIndustry baseline
All files
100%
Baseline
SIU referrals
22% currently referred
Industry avg
Confirmed fraud
9% true fraud rate
cohort study
Caught today
3% caught at intake
~33% of fraud
Owl-detected
8% caught at intake with Owl
~89% of fraud

8%

Fraud surfaced at FNOL

41×

Patterns checked per file

$2.4M

Recovered per 100K files

03Patterns by line of business

Pick a line. See the patterns Owl finds: at FNOL, before the file goes to an examiner.

Each line has its own fraud archetypes. Owl indexes them all and runs every file against the right ones.

OwlVision
What it read
FNOL + recorded statement
14 pp
Read, Contradictions flagged
Medical record (ED + chiro)
184 pp
Read, Pattern-matched
Police report + diagram
8 pp
Read, Geometry-checked
Photos (12 imgs, EXIF)
1 pp
Read, Metadata flagged
Structured extraction
Claim AB-2025-44218, rear-end, claimant 41M, soft-tissue + lumbar, 3 prior soft-tissue claims (ISO), same chiro 2 of 3 priors, same MRI center 2 of 3 priors, same plaintiff attorney all 3 priors
OwlSignal
What it cross-referenced
high
Staged-loss ring, cross-carrier graph

Claimant, treating chiropractor and plaintiff attorney appear together in 14 prior-carrier claims; cluster pattern matches NICB-published staged-loss ring methodology. Refer immediately.

high
EXIF inconsistency

Damage photos timestamped 2025-05-29 (8 days pre-DOL). Photo metadata vs. claimed loss date inconsistent.

mid
Recorded-statement contradiction

Claimant said "no prior back issues" at minute 4:12; ISO ClaimSearch shows 3 prior claims with lumbar treatment. Contradiction documented for EUO.

OwlAssist
What it answered
Examiner asked
Should this go to SIU, and what’s the recovery probability?
Owl answered in 0.7s

Yes: refer to SIU within 24 hours. The pattern is a staged-loss ring with three corroborating signals: cross-carrier provider-attorney cluster, EXIF date pre-loss, and recorded-statement contradiction on prior treatment. Cohort base rate for this signal combination is 71% confirmed-fraud at investigation close. Recovery probability ~$22K average for similar cases. Draft EUO request and SIU referral memo attached.

NICB staged-loss methodologyCross-carrier graph n=14 priorsISO ClaimSearchPhoto EXIF 2025-05-29

How Owl plugs into the SIU workflow. Without changing how SIU runs.

01OWL

FNOL hits the queue

Owl reads the file at intake, runs every fraud-pattern check, scores the file. No SIU referral required to look.

02OWL

Pattern-match, score

Cross-carrier fraud graph, NICB feeds, prior-claim history, document-tampering signals. Score with reasoning.

03OWL

Auto-refer or route

High-confidence patterns auto-refer to SIU. Mid-confidence flag in examiner queue. Low-confidence stays silent.

04INVESTIGATOR

SIU investigator picks up

Loads the file with patterns, priors, base rates and a draft investigation plan. Validates and edits.

05INVESTIGATOR

EUO, surveillance, denial

Owl drafts the EUO outline, the surveillance authorization, the denial framework. Investigator runs the work.

05What changes

Numbers from carriers running Owl in SIU.

MetricWithout OwlWith Owl
Fraud surfaced at FNOL3%8%
Referral conversion (referral → confirmed)38%71%
Cycle time, investigation close94 days41 days
Indemnity savings per 100K files$0.7M$2.4M
False-positive referral rate62%29%
Investigator capacity (files / yr)88224
Source: rolling 12-month average across 5 carriers running Owl in SIU, n = 312K files referred or auto-screened.

Built for the regulatory shape of every state’s SIU statute.

State SIU mandatesNY 11 NYCRR 86, FL 626.9891, CA 1875.20, all 50-state SIU regulations.
NICB reportingMatch-and-refer integrations with ISO ClaimSearch and NICB Foreknowledge.
NAIC Model #680Insurance Fraud Prevention Model Act compliance.
EUO authorityPer-policy + per-state Examination Under Oath authority documented.
HIPAA + state PI lawPHI / PII segmentation, state-private-investigator licensure tracked.
Reg AI / NYDFSBias testing on referral models, model lineage, explainability.
SOC 2 Type IIAnnual audit, continuous control monitoring.
Records retentionPer state, per carrier, per investigation type.
07Integrations

Lives where your SIU team lives.

Claims platforms

Guidewire ClaimCenter, Duck Creek, FINEOS, Origami, Majesco

SIU case management

SAS, NICE Actimize, FRISS, in-house

Industry data

NICB Foreknowledge, ISO ClaimSearch, LexisNexis, CCC

Public records

PACER, state registries, LinkedIn, social media, OFAC, sanctions

Surveillance vendors

BVS, ISG, Hub, Examworks, Frasco

Identity & SSO

Okta, Azure AD, PingFederate, SAML, SCIM

08Get started

Bring us a year of files your SIU never saw. We’ll find the fraud you missed.

Two-week retro pilot. We screen your last 100K closed files against the cross-carrier fraud graph, surface the patterns and stack-rank by recovery probability. You decide what to investigate next.