Risk engine

Risk Analytics

Turn safety and claims data into risk patterns, projections, and decisions.

State

SizeNAICS sectorHigh-confidence onlyExcludes low-exposure establishments from every stat. Key findings state how many were removed.

Benchmark basis

TCIR = all recordable cases. DART = days-away/restricted/transfer, the lost-time measure that drives comp cost.

Data & method

This featured workers'-comp example uses public OSHA establishment injury data as a stand-in for internal claims data. Each establishment's rate is compared with BLS industry incidence rates to produce a synthetic e-mod.

How this works

The OSHA data provides recordable case counts by establishment. BLS industry incidence rates serve as the expected baseline for each industry.

The synthetic e-mod is actual-over-expected and credibility-weighted. 1.0 is industry average; above 1.0 means worse-than-peer claim experience, the same mechanism that raises a real workers'-comp premium.

Proxy caveat: a connected risk program would replace the public proxy with its own safety and claims data.

Workers' compensation · Experience rating

Which establishments are credibly running hot?

Public OSHA injury records, benchmarked against BLS industry rates and blended toward each establishment's own experience by credibility — so a small shop's one bad year doesn't read as a verdict.

Readout

Key findings

Deterministic summary

Rate vs. establishment size

TCR per 100 FTE

The book

Filtered establishments

Sortable table

Industry ranking

Company rate vs. expected baseline

Synthetic e-mod distribution

1.00 = industry average

Loss concentration

P1 mocked Pareto

Frequency vs. severity

DART share of cases