micro1 is recruiting in-house General Counsels to run simulated contract negotiations and redlining exercises that train and evaluate frontier AI models on commercial legal work. You review model redlines against your own, write grading criteria, and feed structured findings back to product and research teams. Observed rates run $90–200/hr on a part-time contract basis.

Summary, eligibility guidance and screening preparation are prepared by Sidequest from the platform's posting. Check the source listing before applying.

What the work actually involves

You will spend most of your time doing what you already do in-house, but instrumented for measurement. A typical task hands you a fact pattern — a vendor MSA with a mutual indemnity, a DPA missing sub-processor flow-down, an NDA with a residuals clause — plus a counterparty position and a client risk posture. You redline it, then compare your markup against the model's. Where the model missed an unlimited liability carve-out, mislabelled a limitation of liability cap as market-standard, or redlined defensively where a GC would have conceded, you write up the miss: what the model did, what a competent GC would have done, and why the difference matters commercially.

The second half of the work is framework building. Contract review is not a single right answer, so the platform needs grading criteria that separate genuine errors from defensible drafting preferences. You will be asked to write rubrics — what counts as a critical miss versus a stylistic quibble, how to score a redline that is legally sound but negotiation-naive — and then apply your own rubric consistently across dozens of samples. Some tasks are multi-turn simulated negotiations where you play counterparty against the model over several rounds.

What the screen looks for

Real transactional depth, not general litigation or generalist practice. Screeners probe MSAs, NDAs and DPAs specifically, and follow up on clause-level mechanics: indemnity scope, liability caps and super-caps, SCC transfer mechanics, IP and feedback licences, SLA credits as exclusive remedy.
In-house judgment. Firm-side drafting experience is welcome but the role is written for people who have owned a risk posture and had to decide what to concede on a deadline.
The ability to explain a redline. Feedback that says "this clause is wrong" is unusable. Feedback that says which party bears which risk under which scenario, and what the market fallback is, is the product.
Consistency. Expect calibration exercises where the same scenario reappears to see whether you grade it the same way twice.

Logistics

Remote, contractor, part-time and largely asynchronous — most experts commit a handful of hours a week around their primary role. Work arrives in batches with turnaround windows rather than fixed shifts, though some calibration sessions and researcher syncs are scheduled live. Volume fluctuates with project cycles; treat this as supplementary rather than a replacement for practice income, and confirm your employer's outside-work and conflicts policy before starting.