What the work actually involves

You build and review evaluation items grounded in real revenue operations practice. A typical task might be a messy Salesforce opportunity export with duplicate accounts and stage-skipping, a pipeline coverage question where the model must reason about weighted vs. unweighted forecast, or a comp plan scenario with accelerators, clawbacks and a mid-year territory change. You write the prompt, produce a reference answer with your reasoning made explicit, and specify what a wrong answer looks like — not just "incorrect" but which assumption broke. On grading shifts, you compare model outputs against that rubric and write short, defensible justifications for the score.

Most of the disagreement in this work is not about arithmetic. It is about whether the model chose a defensible convention: which date defines a closed-won month, whether ramping reps belong in capacity math, whether a renewal counts as new ARR. Reviewers who can name the convention they are applying and why it fits the scenario are the ones whose items survive audit.

What the platform screens for

  • Operator experience, not adjacency. micro1's screen is an AI-led interview with live follow-ups. Claims like "owned forecasting" get probed: which CRM, what cadence, who consumed the number, what happened when it was wrong.
  • Tool specificity. Salesforce or HubSpot object models, reporting layers (Looker, Tableau, Sigma), SQL or spreadsheet modelling, and where the seams are between them.
  • Rubric discipline. Can you convert a judgment call into criteria another reviewer would apply the same way?
  • Written clarity under length limits. Justifications are short by design.

Logistics

Fully remote and async, with no fixed hours; work is claimed from a queue and has turnaround windows rather than shifts. Contributors commonly report 10–20 hours weekly, though volume fluctuates with project starts. Pay is per-hour or per-task depending on the project, and the $50–100 band reflects observed rates rather than a guarantee — placement within it tracks assessed calibration and specialisation. Expect a short paid or unpaid calibration set before full access, and periodic re-review of your graded work.