What the work actually involves

You're producing and grading deck work, not writing opinions about decks. A typical task gives you a brief — a positioning launch, a QBR, a competitive teardown, a pricing narrative — and asks you to build the storyline and slides yourself, or to evaluate what the model produced against a rubric. Evaluation work means reading a generated deck closely and saying precisely where the logic breaks: an action title that states a fact instead of a conclusion, a horizontal flow where slide four doesn't follow from slide three, an exhibit whose chart type fights its own takeaway, an executive summary that summarizes the deck instead of making the argument. Then you rewrite it correctly, because the corrected version is the training signal.

Expect to spend real time on the craft layer PMMs usually take for granted — vertical logic within a slide, the relationship between a title and the body that supports it, whether a build sequence earns each reveal, whether the data on the page actually supports the claim above it. Some tasks are single-slide; some are full client-ready decks with data exhibits.

What the platform screens for

The screen is an AI voice interview. It's looking for whether you've personally built decks that went in front of executives or customers at a company with real velocity, and whether you can articulate why a structure works rather than just producing one on instinct. Expect follow-ups that push on specifics: which deck, what the argument was, what you cut and why, how you knew the storyline held. Vague answers about "driving alignment" collapse quickly under a second question. Named frameworks help only if you can show you use them situationally rather than reciting them.

Logistics

  • Fully remote, fully asynchronous — no standing meetings
  • 5–20 hours per week, with room for more if you want it
  • Work is picked up from a queue on your own schedule
  • $100/hour is the observed rate for this listing, not a guarantee; rates vary by task type and calibration performance
  • Expect an unpaid calibration period at the start where your work is reviewed against the rubric before you're cleared for volume