Proprietary predictive algorithms
Combine classical algorithmic technique with learned components, optimized for accuracy and tightly bound intervals on legal outcomes. This layer produces every published forecast.
The science behind the forecast.
A large language model turns the claim file into structured features.
Mathematical models then use those features alongside millions of resolved cases to predict settlement ranges, escalation probability, and calibrated confidence bands.
Why we splitLLMs and math
Large language models are trained to write plausible text, not to forecast outcomes with a measurable error rate.
A claim forecast needs the opposite: a model whose accuracy is measured on held-out historical cases and published with each prediction.
So we split the work. Language models read the file and turn it into structured features, clauses, parties, dates, medical events. A separate set of mathematical models, trained on millions of resolved cases and calibrated against held-out data, produces the settlement range, the escalation probability, and the comparable cases that anchor each forecast.
One case moves throughpipeline stages.
- 1.
Ingest
- 2.
Embed
- 3.
Predict
- 4.
Calibrate
- 5.
Retrieve
- 6.
Publish
The Model Layers
LLMs as embedders
Language models live inside the predictive framework as feature extractors. They emit normalized representations of clauses, entities, and events so the geometric layer downstream can reason over rich, comparable input rather than raw text.
Geometric machine learning
Cases live as positions in a learned geometric space. Similar cases sit close to one another, and the model uses both the position and the local geometry to forecast outcomes and retrieve comparables.
Signal processing
Time-series methods identify trends and anomalies — escalation signals, jurisdictional drift, social-inflation patterns. This layer surfaces why a file is moving the way it is.
Conformal prediction
Every forecast is wrapped in a calibrated band with a published error rate. A "65% settlement likelihood with a 70–80% band" means that across a thousand claims, the true settlement share lands inside the published bands at the published rate. A "$425K–$550K" damage range is the calibrated 80% interval, not a midpoint with arbitrary width.
Patent-backed proprietary technology
Patent [1]
Symbolic-Neural Legal Forecasting
Neural-symbolic legal outcome prediction — the combination of neural representations of legal text with symbolic reasoning to produce calibrated outcome predictions.
Patent [2]
Legal Intelligence Engine
The legal calculator infrastructure that turns model output into actionable analytics — predictions, ranges, scenarios.
Patent [3]
Claim-Event Graph Network
A graph neural network for tracking parties, claims, events, and dates through complex documents, so downstream models reason about relationships rather than text fragments.
See the math on one of your own files.
Bring one anonymized file and we will run it on a thirty-minute call. You leave with the forecast, the conformal interval, the drivers, the comparables, and answers to the technical questions on the call.