What the work actually involves

You are not using Toast here; you are explaining it in enough detail that a model can reason about it. A typical task might be: write a step-by-step guide for splitting a check four ways after two items were already comped, explain how menu modifiers and item-level pricing interact when a QSR runs a limited-time offer, or read a model's answer about batch settlement timing and mark exactly where it's wrong and why. Other tasks are annotation — labelling authentic menus, order flows, and transaction records so the system learns what a void is versus a refund, or what an 86'd item does downstream.

Expect a mix of written deliverables and occasional live sessions where contributors and researchers refine prompts and task design together. Verbal fluency matters because some calibration happens by call, but the bulk of the output is text.

What the screen looks for

  • Operator depth, not vendor familiarity. micro1's screen pushes on specifics: Toast Tabs versus Toast Now, KDS routing rules, handheld offline mode behaviour, reporting quirks in the Sales Summary. Someone who "has seen Toast at their job" runs out of detail in two follow-ups.
  • Segment awareness. How a fine-dining coursing setup differs from a drive-thru configuration is a recurring probe.
  • Writing that a stranger can follow. Screens frequently ask for a short written sample or a spoken walkthrough judged on precision and ordering of steps.
  • Honesty about edges. Saying "I never handled the multi-location menu hierarchy" scores better than bluffing, because bluffed detail poisons training data.

Logistics and pay

Fully remote, contractor, and largely asynchronous — tasks arrive in a queue with turnaround windows rather than fixed shifts. Contributors commonly report 10–20 hours a week, with some occasional scheduled calls. The posted band of $28–92/hr is wide because micro1 prices by task complexity and demonstrated depth; the upper end is observed for people doing review, calibration, or rubric work rather than first-pass annotation, and nothing about the band is guaranteed. No AI background is expected or asked for.