OuthInstitutional Data.
Made Accessible.
Outh's valuation engine draws from six federal and independent datasets, processed through a four-stage AI pipeline trained on real case outcomes.
Case Classification
Your description is parsed by a fine-tuned language model that maps it to one of 45 sub-categories across employment, personal injury, and education law. The model was trained on attorney intake forms and legal briefs.
Jurisdiction Adjustment
Settlement values vary significantly by state and circuit. Outh applies jurisdiction multipliers derived from regional verdict databases — a wrongful termination in California settles 2.4× higher than the national median.
Evidence Weighting
The strength of your documentation — written policies, medical records, witness statements — is weighted against what courts have historically required to prevail. Cases with stronger evidence profiles receive higher confidence scores.
Probability Distribution
Rather than a single number, Outh outputs a P10 / P50 / P90 settlement range. The P50 is the median expected outcome; P10 and P90 define the 80% confidence interval based on case variance in comparable litigation.
Six Datasets. Updated Quarterly.
Employment discrimination charges, settlements, and consent decrees from 1997—present.
Wage data, earnings distributions, and occupational injury rates used to model economic damages.
FLSA violation settlements, OSHA enforcement actions, and whistleblower award history.
Federal civil case outcomes, jury verdicts, and bench trial decisions across all districts.
Personal injury jury awards and settlement ranges by injury type and jurisdiction.
Title IX and ADA complaint resolutions from the Office for Civil Rights.