What the work involves
This is a head-to-head benchmarking project for a legal AI company. You receive a set of prompts reflecting the kinds of questions a Spanish in-house lawyer actually fields — contract interpretation, commercial terms, supplier and distribution arrangements, corporate housekeeping, compliance touchpoints — and you run each prompt through two LLM platforms. You then compare the two answers side by side and score them against a standardized rubric covering accuracy under Spanish law, correct citation of the Código Civil, Código de Comercio, Ley de Sociedades de Capital and relevant EU-derived rules, completeness, and whether the answer would be safe to act on.
The written component matters as much as the score. Reviewers are expected to name the specific defect — a misstated plazo de prescripción, an invented article number, advice that quietly assumes common-law drafting conventions, a confident answer that ignores mandatory Spanish provisions. "Answer A is better" without a reason is not a usable evaluation.
What the screen looks for
- Verifiable practice history. Five or more years as in-house counsel in Spain, with specifics: sectors, entity types, the matters you personally owned.
- Commercial and contract depth. Expect follow-ups that test whether you know the rules rather than the vocabulary.
- Rubric discipline. Whether you can apply someone else's criteria consistently instead of substituting your own instincts, and whether you can separate a stylistically polished answer from a legally correct one.
- Writing. Concise, specific, in clean English or Spanish as instructed.
Logistics
Fully remote and asynchronous — you work through batches on your own schedule against submission deadlines. The initial commitment is around 10 hours, with continuation likely if calibration and quality hold up. All work sits under NDA. Observed pay for this listing is USD 140–150/hr; rates on Mercor vary by project and are set per engagement, not guaranteed.