The work

This is comparative evaluation, not drafting. You receive a set of prompts covering real Australian legal questions — weighted toward commercial and contract matters — and run each one through two LLM platforms. You then read both outputs closely, judge them against a rubric supplied by the client, and write a concise explanation of your scores. The value you add is legal judgment applied consistently: spotting where a model has cited a repealed provision, conflated state and Commonwealth positions, imported US or UK doctrine, or produced advice that reads fluently but would not survive a partner's review.

Expect the failure modes to be subtle. Models rarely get Australian law spectacularly wrong; they get it plausibly wrong — hedging where the position is settled, asserting where it is genuinely unclear, or missing the practical commercial point an in-house lawyer would flag first. Your written feedback is what makes the score usable to the client, so precision in English matters as much as precision in law.

What the screen looks for

  • Verifiable Australian in-house practice — company, sector, the kinds of matters that actually crossed your desk
  • Working command of Australian commercial and contract law, tested through follow-up rather than self-assessment
  • Ability to apply someone else's rubric faithfully, including when you'd have scored differently on your own criteria
  • Written reasoning that a non-lawyer reviewer could audit

Logistics

Fully remote and asynchronous, done on your own schedule against task deadlines. The initial engagement is around 10 hours with stated potential for more depending on throughput and quality. All work sits under NDA. Pay has been observed at USD 100–120 per hour; rates on Mercor vary by project and are not guaranteed.