Pattern matches
14,207
Identity / network / device recurrences across the book.
Claims Under Continuous Monitoring
52,940
100% of open injury claims
Signals Surfaced · 30d
2,184
412 high severity
Reserve at Risk · Flagged
$51.2M
across 1,284 claims
Indemnity Recovered · YTD
$8.7M
14.1× Carpe Intelligence spend
4-Tier Desk Ladder The Reasoning Engine routes every claim to the lightest desk that safely clears it. Spend scales with signal.
Tier 1 · Auto-Close
38,604 claims
Automated · No Touch
FNOL-scored clean. Zero adjuster effort. Routed straight to payout or denial per policy.
Reserves$214M
Avg cycle14 sec
% of book72.9%
Tier 2 · Standard
11,426 claims
Adjuster-Handled
Normal WC/Auto/GL flow. OIA running quietly in background. Reasoning Engine pre-ranks priority.
Reserves$168M
Avg cycle38 days
% of book21.6%
Tier 3 · Enhanced Review
2,628 claims
OIA + IR Snapshot
Contradiction signals flagged. Snapshot or Standard IR ordered. Adjuster briefed by Reasoning Engine.
Reserves$88M
Avg cycle62 days
% of book4.9%
Tier 4 · SIU Deep Dive
282 claims
Pro+ IR · Carpe Intel
Full analyst workup, image forensics, 16-step OSINT. Court-ready evidence packages.
Reserves$31M
Settlement impact$24.2K avg
% of book0.53%
Claims Reasoning Engine Proprietary AI layer: pattern recognition, contradiction analysis, cross-product learning — 30-day snapshot.
Contradiction flags
3,842
Stated injury vs. observed behavior mismatches. 68% concentrated in WC + Auto BI.
Cross-product signals
891
Signals flowing OIA → IR → Intel — each step compounds the next.
Source: Reasoning Engine telemetry · reviewed 04/22/26 · evidence chain preserved on every flag.
Portfolio ROI · Modeled Annual Impact
Total modeled impact · 95K claims/yr
$30.6M
5–10× documented ROI across deployed carrier base
Severity reduction
$23.6M
Earlier detection → smaller settlements
Litigation reduction
$5.4M
Evidence-driven resolution
Settlement speed
$1.6M
Faster closure when defensible
Carpe spend
$2.0M
100% monitoring · vs. $2.4M internal team at 20%
Live feed · flagged claims
| Claim | Claimant | LOB | Reserve | OIA | Latest signal | Stage | When |
|---|---|---|---|---|---|---|---|
| CLM-2026-48221 | Marcus Chen | WC | $47,500 | 87 | Jet-ski post + GoFundMe — 9 contradictions confirmed by Carpe Intel | Tier 4 · Pro+ IR | 12 min ago |
| CLM-2026-48107 | Danielle Brooks | Auto BI | $32,800 | 79 | CrossFit reel 3 weeks post-DOL; face match 99% | Tier 3 · Standard IR | 1 hr ago |
| CLM-2026-47992 | Dominic Álvarez | WC | $28,400 | 74 | GoFundMe contradicts work restriction; new listing detected | Tier 3 · Monitoring | 2 hr ago |
| CLM-2026-47814 | Rachel Okafor | GL | $19,200 | 64 | Snapshot IR delivered — auto-escalated to Standard by Reasoning Engine | Tier 3 · IR Upgrade | 3 hr ago |
| CLM-2026-47710 | Kevin Alvaro | Auto BI | $41,100 | 81 | Device re-used across 3 prior claims · 2 carriers | Tier 4 · SIU | 5 hr ago |
| CLM-2026-47598 | Jason Park | WC | $22,900 | 58 | LinkedIn shows new coaching role — employment contradiction | Tier 3 · Enhanced | 7 hr ago |