What the work involves

You will spend most of your time doing two things. First, authoring reference deliverables: given a fact pattern — a mid-market manufacturer, a carve-out, a founder-led SaaS target — you produce the slide or exhibit a manager would actually hand a deal team. A normalized EBITDA bridge with adjustment footnotes that survive scrutiny. A net working capital page with the peg, the monthly trend, and the seasonality call-out. A red flag summary that says what the issue is, what it's worth, and what it does to the deal. Second, evaluating model output against that standard: reading an AI-drafted findings deck and marking precisely where it goes wrong — an adjustment that double-counts, a databook summary that reconciles to nothing, a headline that asserts a conclusion the exhibit below it doesn't support.

Written rationale matters as much as the verdict. "This bridge is wrong" is not usable training signal; "the run-rate adjustment for the March headcount reduction overlaps with the already-booked severance accrual, so EBITDA is overstated by roughly the severance amount" is. Expect to write more prose than you did as a manager, and to defend your reasoning when a reviewer disagrees.

What the platform screens for

Ethos runs an AI voice screen before any human conversation. It probes deal-side specifics: how you treated a particular adjustment category, how you'd structure a QoE exhibit for a lender versus a strategic buyer, what you actually put on a page versus in the appendix. Generic diligence vocabulary — "quality of earnings," "proof of cash," "normalizations" — gets follow-ups until you either produce a real example or run out. Slide craft is treated as a technical skill, not a soft one: layout discipline, footnote hygiene, tie-out to source, and the judgment about what belongs on one page.

Logistics

  • Fully remote and asynchronous; no fixed hours, no client calls
  • 5–20 hours per week, with room to take more volume
  • $100/hour is the band observed for this listing, not a guarantee
  • Work arrives as batched tasks with deadlines measured in days, not hours
  • Expect a calibration period where your early submissions get heavy reviewer feedback