What the work actually involves

You will be producing and judging the artifacts a school building runs on — but for a language model. A typical task looks like this: a prompt describes a realistic scenario (a 5th-grade ELA teacher needs a three-week unit plan aligned to specific standards, or a principal needs a Title I family engagement deck for a Tuesday night meeting). You either write the gold-standard reference document yourself, or you review what the model produced and explain precisely where it fails. Deliverables land in the real formats: .docx unit plans, .xlsx pacing guides and data trackers with working formulas, .pptx PD decks with usable slide structure.

The critique is where the value sits. Saying a lesson plan is "weak" earns nothing. Saying that the stated objective is a DOK-1 recall task while the assessment asks for argumentative writing, that the pacing guide allots two days to a standard the district calendar gives eight, or that the progress report narrative uses deficit language that would alarm a parent without giving them an action — that is the work.

What the platform screens for

  • Verifiable K-12 ownership. Four-plus years teaching, coaching, or administering, and direct responsibility for curriculum planning, parent communication, or school improvement documentation. Not adjacent ed-tech or tutoring work.
  • Standards fluency under follow-up. Expect probes on standards-aligned design, vertical alignment, differentiation, and how you would defend an instructional choice to a skeptical department.
  • Actual file craftsmanship. Spreadsheet competence means formulas, conditional formatting, and a tracker a grade-level team could use — not a table typed into Google Sheets.
  • Written precision. Your critique text is the training signal. Vague feedback is unusable.

Logistics

Fully remote and asynchronous; you choose your hours. The stated commitment is 5–20 hours per week, with room to take more if the queue supports it. $80/hour is the observed band for this listing and is not guaranteed — rates vary by task type, batch, and reviewer tier. The interview process includes an AI voice screen, meaning you will speak your answers aloud to an automated interviewer that will ask follow-up questions based on what you say. Work is typically paid per task or per hour logged against reviewed output, so expect an onboarding calibration round before volume opens up.