The research teams behind the best-known AI models come to Mercor for judgment about how office and productivity software is actually used. Your standards and expertise become the rubric that frontier AI is graded against.
About the role
The research teams behind the best-known AI models come to Mercor for judgment about how office and productivity software is actually used. Your standards and expertise become the rubric that frontier AI is graded against.
This is a standing listing for people who work in these tools all day and who want AI training and evaluation work, not a specific job opening. It stays open whether or not a matching project is running this week.
New projects start on short notice, so instead of posting and pulling a listing each time, we keep one open per field and draw from it when one begins.
What the work looks like
We match people in this network to projects like:
- Writing a realistic task from your own work, a document, a model, or a deck, with the starting material and the constraint, plus the reference answer a model's attempt gets graded against
- Grading model-generated documents, spreadsheets, and slides: whether the formula is correct and robust, whether the formatting survives editing, whether the deck actually makes its argument
- Flagging where a model's output opens cleanly and would break in real use, given file conversion, version compatibility, or what happens once someone else edits it
Listings in this field on Mercor have posted at $60 to $70 an hour, set per project by scope and depth.
What we're looking for
- Professional experience working at a high level in office and productivity tools
- Clear written communication, since most of the work involves explaining your reasoning
- Comfort with ambiguous tasks, and willingness to flag unclear instructions
What this covers
This pool spans office and productivity software, including:
- Documents and word processing
- Spreadsheets and modeling
- Presentations and slides
- PDF and document conversion
- Email and collaboration tools
- Office scripting and add-ins
Depth in one tool matters more than passing familiarity with all of them.
How it works …
Official source: work.mercor.com. Details can change; the platform's page is the final word. How we check listings.
