The work

You write and grade. On drafting days you're handed a fact pattern and asked to produce the motion a real client would pay for — a motion to dismiss under 12(b)(6), a summary judgment brief with a statement of undisputed facts, a discovery motion with a meet-and-confer history. On review days you read what a model produced against the same prompt and mark it up: is the standard of review stated correctly, is the cited case still good law, does the argument survive the obvious response from the other side, is the relief requested actually available. Feedback is expected to name the defect and the fix, not register an impression.

A third strand is rubric work. Research teams need grading criteria that two different lawyers apply to the same answer and land in the same place — so you'll be asked to define what separates a 3 from a 4 on citation accuracy, procedural posture, or persuasive structure, and then to apply your own rubric consistently across a batch. This is where litigation experience and evaluation discipline diverge; strong practitioners sometimes grade on instinct and produce scores nobody else can reproduce.

What the screen looks for

  • Documented wins. The three-motion requirement is verified against court orders or docket entries. Have case names, courts, dates, and motion types ready.
  • Federal civil specificity. Expect follow-ups on Rule 12 and Rule 56 mechanics, local rules, page limits, and how you handled an adverse ruling.
  • Feedback that transfers. Can you explain a legal error to a non-lawyer researcher in terms that change how a model is trained?
  • Hallucinated-authority instinct. How you check a citation you don't recognize, and what you do when the model's reasoning is sound but the case is invented.

Logistics

Contractor, part-time, remote, largely asynchronous — work arrives in batches with turnaround windows rather than fixed shifts. Observed rates for this role run $150–300/hr depending on depth of federal motion practice and rubric-authoring experience; rates are set by the platform per project and are not guaranteed. Most contributors commit somewhere in the range of 5–15 hours a week and coordinate with research teams over occasional scheduled calls. The role is designed to sit alongside an active practice, though you should confirm your firm's outside-engagement and conflicts policies before accepting.