What the work looks like

You receive AI-produced artifacts — a DCF in Excel, a credit committee memo, an equity research initiation, a management presentation — and score them against a rubric that covers factual accuracy, analytical rigor, formatting, and presentation quality. In practice this means checking whether the WACC build actually holds together, whether comparables were screened sensibly, whether covenant math ties, whether the deck's charts say what the text claims they say. Most tasks end with written feedback: a short, specific account of what is wrong, how wrong it is, and what a competent analyst would have produced instead.

The reviewing is unglamorous and detail-heavy. Models arrive with plausible-looking outputs and buried circularity, hardcoded plugs, or terminal values that imply nonsense growth rates. Decks arrive with clean layouts and misattributed sources. The value you add is catching the failures that look fine on a first pass.

What the platform screens for

  • Verifiable experience. Five or more years in investment analysis, valuation, or credit — buy-side, sell-side, credit research, transaction advisory, or corporate development. Expect follow-up questions on deals, sectors, and methods you claim.
  • Depth under pressure. Screens push past definitions into judgment: why you'd choose one approach over another, where a method breaks down, what you'd do with incomplete data.
  • Evaluation discipline. Consistency across similar submissions, willingness to distinguish a material error from a stylistic preference, and feedback that a non-expert reviewer could act on.
  • Tooling fluency. Real proficiency in Excel and Google Sheets, plus PowerPoint and Google Slides — you will be judging deck craft, not just prose.

Logistics

Fully remote, hourly, asynchronous. Observed pay for this role band runs $80–120/hr, varying by credentials and task type; rates are set by the platform per engagement and are not guaranteed. Volume fluctuates with project demand — some contributors log a handful of hours weekly, others considerably more during active pushes. There are no fixed shifts, but tasks carry turnaround windows, so reliable weekly availability matters more than any particular schedule.