Criterica Intelligence — production models trained on real court records, not synthetic data
Criterica Intelligence

Know the outcome
before capital moves.

Outcome, settlement, and duration — measured from real court records, governed like production infrastructure, and re-estimated as evidence changes.

NewDuration Intelligence & the research series →
scroll
Decisional Intelligence · New · August 2026

Duration is the variable that converts legal outcomes into financial returns. We made it a managed one.

Time to resolution as a governed distribution: a band with a visible tail, never a date, re-estimated on evidence with the full forecast history preserved. Duration is the first layer of decisional intelligence — the measurement that operational, portfolio, and capital decisions inherit.

DURATION INTELLIGENCE · LIVE
OPERATIONAL INTELLIGENCE
PORTFOLIO INTELLIGENCE
CAPITAL INTELLIGENCE
Each layer inherits the one above it
Start from your seat
Or start with the number: what a point of accuracy is worth on your book →
Where to begin

Built on real outcomes, not confident guesses. Find where you fit.

For institutions pricing legal risk
Run the audit on your own book.

Funders, insurers, law firms, and enterprise legal teams see what the models see in their own portfolio.

Request an Audit
For platforms and partners
Embed the intelligence in your product.

License the scoring layer, the data, or the models. Your interface, our outcomes engine underneath.

Explore Partnership
For strategic acquirers and enterprise partners
Own or integrate the predictive intelligence layer.

A proprietary outcomes corpus and a validated model fleet most of the category cannot replicate — built to sit beneath legal and financial decisions.

Strategic Ownership
Markets
LAW

Structured prediction across civil, commercial, and complex litigation proceedings.

FINANCE

Validated risk intelligence for litigation funders, insurers, and capital allocators.

HEALTHCARE

Regulatory compliance and liability analysis across clinical and institutional settings.

The Problem

Legal capital markets operate on information asymmetry.

The infrastructure to price, underwrite, and manage legal risk has never been built at institutional scale. Criterica is that infrastructure.

01
Capital Deployment Risk

Funders commit capital before structured outcome probability exists. The decision precedes the data by months.

$17B+ committed to US litigation finance (industry estimates) — deployed largely on qualitative case assessment

02
Judicial Opacity

Outcome distributions shift by 20 to 40 percent on judicial assignment alone. Most positions are built without this signal.

Millions of settle-or-litigate decisions are made every year without model-validated outcome intelligence

03
Portfolio Blindness

Correlated case concentration accumulates invisibly at book level. Systemic exposure is only visible after it materializes.

Court records span every major jurisdiction — most of the available signal never reaches underwriters

04
Pricing Asymmetry

Legal risk is priced on actuarial tables and counsel intuition. Case-specific factors remain structurally excluded from the model.

In one deidentified $70M portfolio audit, 44% of active book value carried no risk classification at all

Statistics shown reflect historical or illustrative model outputs derived from real case data. They are not predictions or guarantees of any individual outcome. Litigation results depend on facts, jurisdiction, judge, and counsel, and vary case by case. Model accuracy is subject to selection effects and changing legal dynamics.

The Infrastructure
Pre-Commitment

Validated outcome probability before capital is deployed. Jurisdictional models trained on real court records, not synthetic distributions.

Judicial Intelligence

Judge-level decision history, venue bias, and assignment probability modeled at the circuit, district, and individual level.

Portfolio Analytics

Cross-case correlation, concentration scoring, and stress-test modeling across litigation book positions and time horizons.

What Atlas Includes

One platform. Every function that regulated outcomes intelligence requires.

Platform Modules · Integrated Reports · Dashboard Roles · IC Memo Auto-Generation · Real-Time Signals

01
Underwriting & Case Scoring

IC Memo auto-generation, APPROVE/CONDITIONAL/DECLINE signal, P25–P90 outcome bands.

02
Portfolio Intelligence

Health score, capital velocity, at-risk capital, concentration risk across the full book.

03
Risk Signal Command Center

Score deterioration alerts, escalation queue, stale matter detection, real-time signal feed.

04
Judicial Intelligence

Verified judges. Win rate, reversal rate, ruling patterns — by judge, venue, and circuit.

05
Capital & Facility Management

Facility covenants, UCC calendar, borrowing base compliance, disbursement tracking.

06
Reporting Suite

Integrated reports — portfolio summary, vintage, stress test, LP waterfall, settlement forecast.

Explore the full platform →
Where We Work
See the full market map →

Market-size figures are third-party industry estimates cited for context — they are not Criterica measurements. Portfolio findings referenced on this page derive from deidentified Criterica audits and are labeled as such.

The Moat, in Four Pillars
01
Outcome corpus

Real resolved matters and verified court records — deduplicated, provenance-keyed, and accumulated over years. The part a competitor cannot compress.

02
Model fleet

Jurisdiction- and matter-specific calibrated models, each individually cleared through temporal-holdout promotion gates. Narrow by design, honest by registry.

03
Decision infrastructure

Underwriting, monitoring, reporting, and servicing workflows that put calibrated numbers where decisions actually happen — with humans at the commitment points.

04
Validation loop

Client portfolios and live outcomes continuously test performance. Every resolved case scores the models that predicted it and feeds the next retrain.

Each pillar is replicable in isolation. The moat is the compounding of all four — and the elapsed real-world outcomes there is no shortcut through.

Proof

We ran the audit on a $70M litigation finance book. Here is what it found.

A national pre-settlement fund engaged Criterica to run a Portfolio Intelligence Audit against its active book. The findings were already inside the portfolio. The audit made them legible. Client and counterparties are anonymized, and figures are stated in bands.

~$70M
Active book audited
~44%
Of the book carried no risk score
~$9M
Idle capital surfaced
6 to 18 mo
Earlier warning on failing firms

Built for institutions that require defensible decisions.