Skip to main content
Back to challenges
Challenge

Third-Party Litigation Funding Risk Agent with CrewAI

Legal developments in litigation funding impact insurer loss reserves and settlement strategies. Build a multi-agent crew using CrewAI to orchestrate legal research, docket analysis, and funder tracking across casualty claims portfolios. Achieve an accuracy of 90% in detecting third-party litigation funding involvement in claims.

Machine LearningHosted by Vera
Sign in free to start
Challenge brief

What you are building

The core problem, expected build, and operating context for this challenge.

Build a multi-agent collaborative crew using CrewAI to analyze court filings and detect Third-Party Litigation Funding (TPLF) signatures.

Delivery guide

How work is evaluated

Evaluation

Evaluates TPLF detection accuracy and reserve adjustment calculations.

Datasets

Challenge datasets

Public datasets, plus the private datasets you have access to.

Evaluation rubric

How submissions are scored

These dimensions define what the evaluator checks and which criteria separate a passable run from a strong one.

Dimensions
2 scoring checks
Binary
2 pass or fail dimensions
Ordinal
0 scaled dimensions
Dimension 1

tplf_boolean_check

tplf_present must be a boolean value.

Binary check

This dimension contributes its full weight only when the submission satisfies the requirement. Partial credit is not awarded.

Dimension 2

accuracy

Accuracy of TPLF detection across test cases • target: 0.9 • range: 0.75-1

Binary check

This dimension contributes its full weight only when the submission satisfies the requirement. Partial credit is not awarded.

Learning goals

What you should walk away with

  • Orchestrate multi-agent roles using CrewAI (Researcher, Legal Analyst, Reserve Risk Manager)

  • Build sequential and hierarchical crew task workflows

  • Detect Third-Party Litigation Funding (TPLF) indicators in legal dockets

  • Achieve 90% classification accuracy on TPLF involvement flags

Resources and assets

Reference links and supporting material

Dataset notes

Dataset of 600 civil litigation docket summaries from state and federal courts.

Challenge setup

Evaluates TPLF detection accuracy and reserve adjustment calculations.

Challenge input

JSON containing docket_text, law_firm_name, and motion_history

CrewAI

Enables role-playing agent delegation for multi-stage legal docket evaluation.

Evaluated output

JSON with tplf_present (boolean), funding_entity (string or null), and reserve_increase_pct (float)

Evaluation checks
  • tplf_present must be a boolean value.
  • Accuracy of TPLF detection across test cases • target: 0.9 • range: 0.75-1
Configured evaluation evidence
  • Benchmark: TPLFDetectionBench
  • Accuracy target: 0.9
  • 1 public reference case
Configured runtime
  • Python execution harness
  • Python sandbox (unavailable on Versalist)

Summary shows the first task, up to three tools, and up to three evidence items. Open the technical recipe to view all configured items.

View technical recipe

Configured tools

Action Space
  • CrewAI · Required
  • crewAI · Optional
  • BentoML · Optional

Evaluation contract

  • tplf_boolean_check · Weight 1
  • accuracy · Weight 1

Metrics

  • accuracy · Target 0.9

Benchmarks

  • TPLFDetectionBench
  • general

Reference cases

  • 1 public reference case

Workflow tasks

  1. Task 1

    Input: JSON containing docket_text, law_firm_name, and motion_history

    Output: JSON with tplf_present (boolean), funding_entity (string or null), and reserve_increase_pct (float)

    Execution harness: Python

Evaluation tools

  • CrewAI · Agent Framework

Recipe freeze marker

No recipe freeze time is recorded.

Frequently Asked Questions about Third-Party Litigation Funding Risk Agent with CrewAI