What the work involves

You will run simulated negotiations and redlining exercises on ISDA Master Agreements and related documentation — Schedules, Credit Support Annexes, confirmations — and then assess how a frontier model handles the same material. A typical task pairs a derivatives scenario with one or more model outputs: you judge whether the model correctly identified the Event of Default versus Termination Event distinction, whether it applied close-out netting mechanics properly, whether a proposed redline actually shifts risk the way its rationale claims, and whether the drafting would survive a counterparty's review. Beyond pass/fail judgments, you are asked to write the grading criteria themselves — objective rubrics that another reviewer could apply to the same output and reach the same conclusion.

The second half of the job is translation. A model that mishandles Additional Termination Events needs feedback more specific than "incorrect" — it needs an explanation of what the provision does, why the output is commercially wrong, and what a buy-side versus sell-side negotiator would actually have pushed for. Product and research teams use that written reasoning to refine guidelines, so the quality of your prose matters as much as the correctness of your call.

What the screen looks for

micro1's screening is AI-led and conversational, and it probes depth rather than résumé coverage. Expect follow-ups that keep drilling on a single provision until you either demonstrate working knowledge or run out of it: what a Section 5(a)(vi) cross-default threshold does in practice, how Automatic Early Termination interacts with particular insolvency regimes, why a client would accept a one-way Break Clause. Claims about years of experience should be attached to specific documents you negotiated and specific positions you took. Evaluation judgment is tested separately — the screen wants to know whether you can distinguish a model answer that is wrong from one that is merely not how you would have drafted it.

Logistics

  • Remote, contractor, part-time; work is asynchronous and task-queued rather than scheduled
  • Compensation is task-based, with average handling time around 3.5 hours per task; effective hourly depends on tasks completed and signed off
  • Pay band: $90–130/hr (observed, not guaranteed)
  • Minimum three years of hands-on derivatives documentation experience, including ISDA negotiation
  • Prior legal tech or AI training exposure is preferred, not required