The work

You receive AI-generated artifacts — a Series B pitch deck, a lender presentation, a CIM section, a management-case model tab, a one-page teaser — plus a rubric and the prompt the model was given. Your job is to grade the output the way a deal lead would mark up an associate's draft: is the equity story coherent, are the KPIs the ones investors in this sector actually underwrite, do the numbers on the traction slide reconcile to the model, is the cap table math right, does the chart actually support the headline above it? Written feedback matters as much as the score. Ratings tell the training pipeline which output won; your comments explain what a credible fundraising professional would have done instead, in terms specific enough to act on.

Expect a mix of tasks: side-by-side comparisons of two decks, single-artifact scoring against a rubric, and occasional rewrites or gold-standard exemplars. Aesthetic and presentation judgment counts here more than in most finance evaluation work — alignment, hierarchy, chart type selection, and slide density are graded, not just content accuracy.

What the screen looks for

  • Verifiable depth. Which stages and instruments you've raised or advised on, deck counts, sectors, whether you sat on the banker side, the sponsor side, or in-house at the company.
  • Specific failure modes. Reviewers who can name the common tells of a synthetic pitchbook — invented comps, generic TAM stacks, market-size math that doesn't survive a second question — score better than reviewers who describe quality in adjectives.
  • Rubric discipline. Whether you can separate "I would have written it differently" from "this is wrong," and hold a consistent standard across dozens of similar artifacts.
  • Tooling fluency. Real working proficiency in Google Slides and PowerPoint, plus Excel/Sheets, since you will be inspecting and sometimes editing files.

Logistics

Fully remote, contractor, hourly, invoiced through Mercor. Work is asynchronous with task-level or weekly deadlines; most evaluators commit 10–20 hours a week and choose their own hours. Volume varies by project cycle — periods of steady queues followed by gaps are normal. Pay bands are as observed on the platform for this role and are not guaranteed; final rates depend on screening outcome and project.