INSURANCE INTELLIGENCE
Generally availableRoute every claim narrative to the right queue - and the staged loss to SIU.
200 claims-intelligence labels across 40 languages - claim types, staged fraud, exaggeration, coverage disputes - each returned with the configuration ID and receipt you need to reproduce the call. No generative model is used to produce classifications.
200 labels · 4 signal groups · 40 languages · p95 target <500ms · No generative model is used to produce classifications
Re-run with the same pinned version → identical result
What this analyzer does and doesn't do
- Detect 200 labeled claim, fraud, and dispute signals in claim text
- Return confidence-ranked signals your workflow engine can route on
- Cover 40 languages
- ✕Access claim systems, policy records, or damage photos
- ✕Determine whether a claim is actually fraudulent
- ✕Make payout, denial, or settlement decisions
- ✕Replace your adjusters or SIU investigators
Sentiqra returns structured signals. Your policy, reviewers, and models make the decision. See the full boundary statement →
WHAT YOU'LL DETECT
Not "suspicious claim: 0.91". The specific pattern - with the label your SIU playbook already knows.
Staged & Deliberate Fraud
fraud_staged_accident
“The other car brake-checked me - I had no time to stop” → very high
Claim Exaggeration & Inflation
exag_preexisting_damage
“All of this damage is from this one incident” → very high
Coverage Disputes & Litigation
legal_bad_faith
“Your denial is unreasonable - my attorney will be in touch” → very high
Organized Fraud Rings
ring_medical_mill
Detected - example withheld
Documentation & Identity Fraud
doc_forged_receipt
Detected - example withheld
200 indicators across 4 signal groups - covering 40 languages.
Same deterministic product contract as every Sentiqra analyzer: pinned configuration, signed receipt, replayable results. Per-language benchmark publication is pending for this analyzer - evaluate it on your own labeled workload; the published methodology applies when results ship.
WHERE IT GETS USED
Three jobs this analyzer does well.
First-notice-of-loss claims triage
Every FNOL narrative runs through Insurance Intelligence at intake. claim_type_auto, claim_type_property, and claim_injury_severity signals route the claim to the right line-of-business queue with severity already scored - no adjuster reads a narrative just to decide who should read it.
Fit signals: high FNOL volume · multi-line books · multilingual policyholders
SIU referral screening
fraud_staged_accident, exag_inflated_injury, and ring_medical_mill hits assemble the SIU queue automatically. The signed receipt on every referral shows investigators exactly which indicators fired - replayable while the configuration remains supported - including when the case reaches litigation.
Fit signals: SIU backlog · litigation exposure · audit trail required
Coverage-dispute early warning
denial_not_covered disputes and legal_bad_faith language surface in policyholder correspondence before they become regulator complaints. Customer advocacy gets the case while it's still a service problem, not a legal one.
Fit signals: regulatory exposure · complaint escalation costs · retention pressure
Send text. Get ranked labels + a receipt you can replay.
POST /v1/analyze
Authorization: Bearer sq_live_your_api_key_here
{
"text": "The other driver cut in front and slammed on the brakes - I had no time to stop. My neck still hurts from the last accident too.",
"language": "en",
"analyzers": ["insurance"]
}{
"language": "en",
"signals": [
{ "analyzer": "insurance", "code": "claim_fraud", "label": "Claim Fraud", "score": 0.9, "confidence": "very_high", "level": "category" },
{ "analyzer": "insurance", "code": "fraud_staged_accident", "label": "Staged or deliberate auto accident", "score": 0.9, "confidence": "very_high", "level": "label", "parent": "claim_fraud" },
{ "analyzer": "insurance", "code": "fraud_swoop_squat", "label": "Swoop and squat or forced collision scheme", "score": 0.9, "confidence": "very_high", "level": "label", "parent": "claim_fraud" }
],
"meta": { "analyzers": [{ "name": "insurance", "status": "completed", "signal_count": 3, "billed": true }], "partial": false },
"usage": { "analyzer_runs": 1, "billed": true },
"verification": {
"configuration": "sqcfg_GFAWTERR",
"fingerprint": "fp_8a72...19f",
"input_commitment": "ic_31bd...a42",
"signature": "sig_ed25519_4c1f...e2a"
}
}READING THE OUTPUT
What a Insurance Intelligence result tells you - and what it doesn't.
Intended task
Classify claim narratives into documented claim types and fraud-pattern signal groups - for routing and SIU review.
Input requirements
text- the only field analyzed; up to 4,096 Unicode code points. Over the limit is rejected withinput_too_long, never truncated.language- must be a language this analyzer supports. Unsupported codes are rejected with400 unsupported_language, never analyzed as a different language.analyzers- include"insurance"; up to 2 analyzers per request, one billed run each.- No context beyond
text- no user IDs, history, profiles, or attachments are read or inferred. - Analyzes the narrative text only - policy terms, claim history, and evidence files are not inputs.
Ordinary positive
“My parked car was hit overnight. The other driver left a note with their plate number and insurer.”
Safe inference: The narrative classifies as an auto claim with a third-party element - route it to the auto queue with subrogation flagged.
Confusing negative · Quotation
“The claimant's own witness said "we agreed to split the payout" - see the attached statement.”
What happens: Fraud-pattern signals can fire on the quoted statement. The signal flags that the language matches a documented fraud pattern - it does not weigh whose words they are or whether the statement is credible.
Why it's tricky: Claim files are full of quoted speech - witness statements, adjuster notes, reported conversations. A fraud-pattern signal on a narrative is a review trigger, never a finding: route it to your SIU, and let investigators weigh the evidence.
What you can safely infer
- Which documented claim types and fraud-pattern groups the narrative matched.
- That the same narrative on the same pinned configuration classifies identically - the receipt supports dispute review.
What you must not infer
- ✕That fraud occurred or that the claimant lied - fraud signals route review, they do not determine wrongdoing.
- ✕Coverage, liability, or payout decisions - those stay with your adjusters and policy terms.
- ✕That a claim with no fraud signal is verified legitimate - a missing signal is "not detected," never "cleared."
Boundary: Fraud signals are not proof of wrongdoing. They flag language consistent with documented fraud patterns so a reviewer looks - the determination stays with your SIU and your process.
Insurance Intelligence is on every plan. Even Free.
Free
5,000 runs/mo
Build
$149/mo · 50,000 runs/mo
Growth
$349/mo · 150,000 runs/mo
Production
$799/mo · 500,000 runs/mo
Scale
$2,999/mo · 2,500,000 runs/mo
Enterprise
By agreement · SOC 2 Type II in progress
Deterministic by construction, provable by receipt.
A signed verification receipt - configuration ID (sqcfg_...), result fingerprint (fp2_...), input commitment (ic3_...), and Ed25519 signature (sig_ed25519_...) - ships with every successful analysis result. Pin it to the stored decision and a later replay under the same supported configuration and output profile yields the identical canonical result - the difference between "the model said so" and evidence.
Where it fits in your pipeline
Often paired with: Fraud & Scam for premium and payment fraud, Customer Service Intelligence for policyholder support signals, Emotion Analysis for distress and vulnerability scoring.
Used in: Customer Support Intelligence and Multi-Domain Analysis
Try it on your hardest claim narratives.
5,000 analyzer runs/month · No card required
Analyzer FAQ
New to Sentiqra? It's a deterministic text-analysis API - you send text, you get labeled, confidence-scored signals plus a replayable receipt. Insurance Intelligence is the insurance risk & claims signal detection analyzer.
What does Insurance Intelligence detect?
Detect claim events, policy-risk signals, coverage disputes, fraud indicators, and escalation language in multilingual insurance text.
Which languages does it cover?
40 supported languages. Per-language benchmark publication is pending for this analyzer - evaluate it on your own labeled workload before relying on a specific language. Coverage is per-analyzer, not platform-wide.
How do I call it?
POST /v1/analyze with "analyzers": ["insurance"]. Same endpoint, auth, and response shape as every other analyzer - you can pair it with one more analyzer in the same request.
What does a call cost?
One analyzer run per call, billed against your plan's included runs (overage per plan). All GA analyzers are on every plan, including Free - see pricing.
How do I evaluate it on my data?
Label a few hundred of your own examples, run them through the free tier, and compare our signals to your labels. The evaluation playbook walks through building that test set - it's the same method we use internally.