What the work involves
You'll spend most of your time on slide-level judgment. A typical task gives you a prompt and context — say, a sell-side pitch page, a portfolio company board update, or a three-page executive summary — and either asks you to produce the deck a good analyst would produce, or hands you a model-generated version to grade and rewrite. The recurring themes are storyline development, horizontal and vertical logic, action-titled slides, exhibit design that actually supports the title, and the difference between a deck that reads internally and one that goes to a client.
Critiques are written, detailed, and expected to name the flaw rather than gesture at it: the title asserts a conclusion the chart doesn't show, the three buckets on the page overlap, the executive summary buries the ask on page two. Vague feedback ("could be clearer") is the single most common reason contributor work gets rejected on this kind of project.
What the platform screens for
Ethos runs an AI voice screen before any task work. It probes where you worked and on what — deal types, deck types, who the audience was, what your review cycle looked like. Expect follow-ups that test whether you've actually built these pages or only seen them: how you'd structure a page whose data doesn't support the intended message, when a waterfall beats a bridge, what you'd cut from a twelve-page exec summary. Generic consulting-speak about "MECE" without a worked example tends to fall apart under two follow-ups.
Logistics
- Fully remote, fully async — no standing meetings, no fixed hours
- 5–20 hours per week, more if you want it; work is claimed from a queue
- Rate observed at $100/hour; pay bands on these labs move and are not guaranteed
- Expect to work in PowerPoint and to deliver both files and written rationale
- Engagements are typically project-length and can extend or wind down with lab demand