What the work involves

You will spend most of your time doing two things. First, redlining and negotiating contracts under structured prompts — an MSA with a lopsided limitation of liability, an NDA with a residuals clause, a DPA missing sub-processor flow-down — so the model has expert-quality reference work to learn from. Second, grading what the model produces: did it catch the uncapped indemnity, did it invent a statutory obligation, did its proposed markup actually move the risk allocation or just reword the clause. Feedback has to say why the redline is wrong and what a competent in-house counsel would have done instead, in language a research team without a JD can act on.

A meaningful part of the role is framework-building rather than line-level review. micro1 asks contributors to write grading rubrics — what separates a 3 from a 5 on a fallback position, how to score a model that spots the right risk but misstates the governing standard — so that multiple attorneys scoring the same output land in the same place. That work is closer to drafting a negotiation playbook than to answering questions.

What the screen looks for

The screening is AI-led and follow-up heavy. It probes whether your transactional experience is genuinely in-house and genuinely tech — expect to be pushed on specific clauses you have fought over, which positions you conceded and why, and how your company's risk posture shaped your fallbacks. Outside-counsel-only or litigation backgrounds tend to surface quickly under that questioning. It also tests evaluation temperament: whether you can separate "not how I would draft it" from "legally wrong or commercially unacceptable," and whether you can hold a standard consistently across dozens of samples rather than re-litigating each one.

Logistics

  • Remote, contractor, part-time; contributors commonly report 10–20 hours a week with flexible scheduling.
  • Largely asynchronous, with occasional calibration sessions or syncs with product and research staff.
  • Observed pay band is $85–105/hr, typically set by depth of transactional experience and not guaranteed for any given engagement.
  • Work is delivered through the platform's task interface; volume can fluctuate between project phases.