AI Litigation Outcome Prediction Engineer

Selvix Works

Remote · $160k – $200k / year · Full-time · Posted 4 months ago

This role is no longer accepting applications

Don't miss the next role like this

Create a free profile with your skills. We match you to new AI roles as they open, and companies hiring can contact you directly.

Create free profile, 2 minutes

Similar open roles you can apply to now:

See all open AI jobs

This role is no longer accepting applications

Don't miss the next role like this

Create a free profile with your skills. We match you to new AI roles as they open, and companies hiring can contact you directly.

Create free profile, 2 minutes

Similar open roles you can apply to now:

See all open AI jobs

About the role

A partner at an Am Law 100 firm decides whether to take a contingency case partly by gut feel, because predicting litigation outcomes is genuinely hard — judge variability, jurisdiction quirks, opposing counsel tendencies, case fact patterns that superficially resemble precedents but diverge in key ways. Our platform gives litigators a data-driven second opinion. We've indexed 4.2 million federal and state court decisions with structured metadata on judges, counsel, case outcomes, and procedural history. The prediction models we've built achieve 71% accuracy on held-out cases in federal court. The target is 78%. The bottleneck is better case fact representation — our current feature engineering doesn't capture the legal reasoning patterns that actually drive outcomes. You'll own model development for the litigation prediction product. Lawyers will scrutinize your methodology, so explainability matters as much as accuracy.

Fully remote.

Responsibilities

Redesign case fact representation using legal domain language models
Build judge and jurisdiction-specific calibration for outcome predictions
Implement uncertainty quantification for model confidence signaling
Create explainability features that surface the key factors driving predictions
Design evaluation frameworks that measure accuracy on legally meaningful subgroups
Present methodology to legal professionals and withstand rigorous questioning

Requirements

5+ years in NLP or ML engineering with complex text analysis experience
Experience with legal, biomedical, or other specialized domain NLP
Strong PyTorch and statistical computing skills
Understanding of calibration, uncertainty estimation, and model explainability
Ability to communicate technical methodology to demanding, non-technical audiences