What the work involves
You'll spend most of your time in one of three modes. First, authoring reference material: given a fictional or anonymized startup brief, you build the deck a founder should be walking into a partner meeting with — narrative arc, traction slide, market sizing exhibit, financial summary, the works. Second, evaluating model output: the AI produces a deck or a single slide, and you grade it against a rubric and write the reasoning for why a Series A investor would or wouldn't lean in. Third, repair: you take a near-miss deck and rewrite it, documenting each change so the model learns the delta, not just the destination.
Typical task surfaces include seed-to-Series-C pitch narratives, traction and metrics slides, TAM/SAM/SOM exhibits and the sourcing behind them, financial summary pages, data room indexes, and investor Q&A prep documents. Written justification matters as much as the artifact — a great slide with a two-line explanation is worth less here than a good slide with a paragraph on why the alternative framings fail.
What the platform screens for
Ethos runs an AI voice screen before any project work. It probes for specifics that can't be bluffed: which rounds you worked on, what stage, what the deck actually had to overcome, how you handled a weak metric. Expect follow-ups that push one layer past your first answer. The screen also tests evaluation judgment — whether you can look at mediocre output and articulate why it's mediocre in terms an investor would recognize, rather than reacting to visual polish. Slide craftsmanship in PowerPoint or Google Slides is assumed; the differentiator is whether your reasoning about investor psychology holds up under questioning.
Logistics
- Fully remote, asynchronous, no fixed hours
- 5–20 hours per week, with more available if you want it
- Observed rate around $80/hour; not guaranteed and varies by task type and project
- Work is claimed from a queue; consistency week to week matters more than volume in any single week