What the work involves

You build self-contained litigation "worlds" — a fact pattern, a procedural posture, a document set, and a task an AI system is asked to perform. That might be drafting a motion to compel against a deficient production, assessing privilege and work-product claims across a custodian set, sequencing depositions in a multi-defendant commercial dispute, or preparing a section of a trial memorandum. For each scenario you write the reference output a competent senior litigator would produce, then author the rubric that explains what separates that answer from a plausible-sounding one.

Scenario coverage spans pre-trial discovery, motion practice, trial strategy, settlement negotiation, and arbitration procedure, plus class and complex litigation and appellate work. Many tasks involve realistic tooling context — e-discovery platforms such as Relativity or Everlaw, litigation management systems, deposition and trial prep software — because the point is to test judgment in the conditions litigators actually operate under.

What the platform screens for

  • Verifiable matter ownership. Screens probe specific cases you ran or contributed to: posture, your role, what you filed, how it resolved. Generalities read as thin.
  • Procedural precision. Expect follow-ups on rule mechanics — Rule 26(b)(1) proportionality, Rule 37 sanctions posture, FRE 702 and 803 applications, or CPR Part 31 disclosure and ICC Article 22 case management on the international track.
  • Rubric instinct. Can you articulate, in writing, why one answer is wrong in a way a non-lawyer grader could apply consistently? This matters more than credential prestige.
  • Discrimination between recall and judgment. The stated goal is tasks that defeat bar-exam-level pattern matching. You will be asked how you would do that.

Logistics and pay

Fully remote and asynchronous, contractor engagement, project-based. Contributors typically commit 10–20 hours per week with flexibility on scheduling, though throughput expectations are real and deadlines are per-batch. Observed rates for this listing sit at $90–100/hr; actual offers depend on track, credential, and calibration performance, and are not guaranteed. Onboarding usually includes a paid or unpaid calibration task before sustained volume is assigned.