What the work involves

You will be handed documents of the kind you already deal with professionally — RFPs and RFI responses, procurement solicitations, grant narratives, agency policy memos, contract clauses, capture plans, compliance checklists, public-sector pricing sheets — and asked to judge them. Sometimes the artifact is model-generated and you assess whether it would survive contact with a real contracting officer. Sometimes you compare two model outputs and justify a preference. Sometimes you write the reference answer yourself, so the model has a defensible target.

Feedback is structured, not impressionistic. Expect rubrics with named dimensions (factual accuracy, regulatory grounding, completeness against the solicitation, tone appropriate to the audience) and a written rationale that cites the specific line or omission that drove your score. Reviewers who write "weak response" without pointing at the clause it missed get sent back for revision.

What the screen looks for

  • Verifiable domain specifics: which agencies, which vehicles, which stage of the procurement cycle you actually touched, and what you personally produced.
  • Depth under follow-up. The AI screener will push a second and third layer on any claim — naming FAR without being able to describe how a part shaped a document you wrote is a common failure point.
  • Evaluation judgment: whether you can separate a confidently written answer from a correct one, and whether you can hold a rubric line rather than substituting personal preference.
  • Real fluency in Excel, PowerPoint, and Word, since much of the source material and deliverable formatting lives there.

Logistics

Fully remote and largely asynchronous, with occasional calibration calls. Most contributors are paid hourly against tracked, reviewed work and choose their own hours; volume moves with project cycles, so treat this as a variable-load engagement rather than a steady weekly commitment. Eligibility is limited to candidates based in the United States, Europe, the UK, or Canada. Rates in the $80–160/hr band are as observed on the platform, not a guarantee — placement depends on domain scarcity and calibration performance.