What the work involves

You'll spend most sessions doing one of two things. The first is authoring: given a scenario — a three-workstream ERP rollout, a post-merger integration, a regulatory remediation program — you build the real deliverable. A program charter with scope boundaries that actually exclude something. A RAID log where risks carry owners, mitigation, and a distinguishing line between a risk and an issue. A resource capacity model where the FTE math reconciles across roles and months. A steering committee deck that leads with the decision being asked for, not a timeline slide.

The second is evaluation: you read what the model produced against the same brief and judge it. That means catching the capacity spreadsheet that allocates 1.4 FTE of a person who is 0.5 allocated elsewhere, the status deck that reports green on a milestone whose dependency slipped two weeks ago, the stage-gate packet missing exit criteria entirely. Written rationale matters as much as the score — the lab is training on your reasoning, so "the RAID log conflates assumptions with dependencies and gives no owner for three of seven risks" is useful in a way that "this is weak" is not.

What the platform screens for

Ethos runs an AI voice screen before any work is assigned. It probes for direct ownership rather than adjacency: whether you wrote the charter or reviewed someone else's, whether you presented at steering committee or prepared slides for the person who did. Expect follow-up questions on specific artifacts — how you structured a dependency tracker across workstreams, how you handled a resource conflict between two program leads, what your stage-gate criteria looked like. Vague methodology talk gets probed until it either produces specifics or doesn't.

Spreadsheet fluency is tested concretely. You should be comfortable explaining how you'd model capacity with ramp-up curves, partial allocations, and leave, and what breaks when a program adds a workstream mid-flight.

Logistics

  • Fully remote, fully asynchronous — no standing meetings, no fixed hours
  • 5–20 hours per week, self-scheduled, with room for more if you want it
  • $70/hour is the observed rate for this listing; it is not guaranteed and can vary by task type and volume
  • Work arrives in batches; pace tends to be uneven, so people who can absorb a heavier week occasionally do best