What the work involves

You receive AI-generated work products — a marked-up services agreement, a diligence summary of a data room, an issues list, a closing checklist, a slide deck summarizing contract risk for a deal team — and grade them against a rubric. The judgment being captured is yours as a practitioner: whether an indemnity carve-out was actually drafted, whether the model invented a defined term, whether a diligence flag is material or noise, whether a redline preserves the client's negotiating position or quietly concedes it. Written feedback matters as much as the score; the notes are training signal, so vague comments like "not thorough enough" are less useful than "missed that the limitation of liability cap excludes the confidentiality obligation, which is the point of the clause."

Tasks span transactional and commercial contracting work: NDAs, MSAs and SOWs, SaaS and licensing terms, DPAs, employment agreements, M&A diligence and reps/warranties review, and playbook-driven redlining. Deck and spreadsheet quality counts — presentation errors, broken formatting, mislabeled exhibits, and unreadable slides are in scope, which is why fluency in Google Slides and PowerPoint is a stated requirement rather than a nicety.

What the platform screens for

  • Verifiable practice history. Five-plus years doing contracts, diligence, or redlines — in-house, at a firm, or in a contract management or legal ops function. Expect follow-ups on deal types, counterparty postures, and what you personally drafted versus reviewed.
  • Depth under pressure. The AI interviewer will push on specifics: how you'd handle a mutual indemnity request, what you check first in a target's customer contracts, where a limitation-of-liability clause typically breaks.
  • Evaluation judgment. Can you separate a confident-sounding but wrong answer from a defensible one, and can you rank errors by severity rather than listing everything you noticed?
  • Written clarity in English, since feedback is the deliverable.

Logistics

Fully remote and asynchronous. Work is drawn from a task queue, so volume fluctuates — some weeks offer 10–20 hours, others less, and consistent quality scores tend to correlate with steadier allocation. No fixed hours or client calls. You are not giving legal advice and there is no attorney–client relationship; confirm your own bar or employer conflict rules before taking on outside review work. Pay is hourly, invoiced through the platform; the $80–120 band reflects observed rates and is not guaranteed.