What the work involves
Project Rachel publishes legal-benchmark tasks and exposes them to public commentary. Your job starts after that: someone has argued that a prompt is ambiguous, that a rubric line item is wrong on the law, or that the attached documents don't support the graded answer. You adjudicate that challenge — sustain it, overrule it, or narrow it — with reasoning a non-specialist reviewer can follow.
The second half is structural QC. That means checking the prompt against the document set it ships with, confirming every rubric criterion is actually answerable from those documents, catching exhibits that are referenced but missing, and flagging rubric items that reward a conclusion the record doesn't support. Task packages typically run 10–100 source documents: purchase and sale agreements with their schedules and disclosure packages, lease abstracts and rent rolls, title commitments with survey and exception review, and closing checklists with the associated deliverables.
What the platform screens for
- Real, transactional real estate practice — not adjacent litigation or generalist corporate work. Screeners probe for the specific documents you've drafted, negotiated, or reviewed, and how many.
- Willingness to say a rubric is wrong and explain why, including when the challenge came from someone arguing in bad faith or citing the wrong jurisdiction.
- Comfort with jurisdictional variance: title practice, transfer taxes, recording requirements and lease conventions differ enough that a rubric written for one state can be defensible and still wrong elsewhere.
- Stamina for document-set review. A 90-document closing package is the normal case, not the hard one.
Logistics
Fully remote and asynchronous. Reviewers pick up task packages from a queue and return adjudications and QC notes; there are no standing meetings, though calibration threads and disagreement escalations run in writing. Expect to commit to a stated weekly volume rather than fixed hours — many reviewers work 10–20 hours a week alongside practice. The $120/hr figure is as observed on this listing and is not a guarantee; rates on Mercor vary by seniority, task type, and calibration performance.