The work
You draft multiple-choice questions from your own knowledge of Thai law — no scraping from bar exam banks or textbooks. Each item needs a clear stem, ten answer options where the distractors are genuinely tempting to someone with partial knowledge, one defensible correct answer, and a worked solution that cites the governing provision or precedent (Civil and Commercial Code, Penal Code, Land Code, Revenue Code, Labour Protection Act, Constitutional Court or Supreme Court rulings, as relevant).
The purpose is adversarial: the questions are meant to find the edges of what a language model actually understands about Thai law, so items that a competent model answers correctly on the first pass have limited value. Strong contributors tend to work in areas where Thai doctrine diverges from common-law or generic civil-law intuition — nominee shareholding rules, foreign land ownership restrictions, lèse-majesté procedure, sakdina-era property concepts still embedded in current statutes, Islamic family law in the southern border provinces.
What the screening looks for
- Verifiable credentials: a Thai LL.B. or foreign equivalent, and ideally two or more years in practice, judicial service, prosecution, or academia.
- Native-level Thai, since primary sources and much of the reasoning are in Thai.
- Precision under follow-up. Expect the screen to press on a claim you make and ask for the section number, the year of amendment, or how a court has actually applied it.
- Judgment about difficulty calibration — whether you can tell a hard question from a merely obscure one.
Logistics
Fully remote and asynchronous. There is no fixed schedule; contributors typically commit somewhere between five and twenty hours a week and are paid hourly at the observed band above, which is not guaranteed and may vary by batch and assessed seniority. Output passes through reviewer QA, and items rejected for ambiguity, duplicated reasoning, or unsupported solutions may need rework.