Expert Project Manager

$50–60/hrrate as listed

Run day-to-day operations for data projects that train LLMs on web browsing tasks — tracking annotator throughput and quality in spreadsheets, reviewing work samples, and keeping contributors unblocked. Mercor lists this at $50–60/hr as observed on the platform; hours and duration vary by project.

Summary, eligibility guidance and screening preparation are prepared by Sidequest from the platform's posting. Check the source listing before applying.

What the work actually looks like

This is an operations seat on a data project, not a strategy seat. On a typical day you're pulling annotator throughput and quality numbers into Google Sheets, spotting the contributors whose accuracy is drifting, sampling their recent submissions to figure out why, and writing the message that corrects it. You'll field contributor questions about ambiguous task instructions, escalate genuine spec gaps to the research or client side, and keep a running picture of whether the project will hit its volume target on schedule.

The subject matter is LLM browsing — tasks where a model navigates the live web to answer questions or complete multi-step goals. You don't need to have built one, but you do need to be able to read a task rubric closely enough to tell a real annotator error from a rubric ambiguity, because that distinction drives most of your decisions.

What the screen looks for

Concrete ownership stories. Expect follow-ups on scale: how many contributors, what volume, what the quality metric actually was and how it moved.
Spreadsheet fluency under questioning. Not "proficient in Excel" — how you'd structure a tracker, what formulas or pivots you'd reach for, how you'd catch a data-entry problem before it reaches a report.
Evaluation judgment. Scenarios where two annotators disagree, or where a top performer's numbers look too good, and what you'd do next.
Written communication. Much of the role is asynchronous messaging to contributors who may be in other time zones; the screen will probe tone and clarity when delivering corrective feedback.

Logistics

Remote and largely asynchronous, though PM roles on Mercor tend to carry more real-time overlap expectations than annotation work — some projects want a few hours of predictable daily availability for contributor escalations. Engagements are commonly structured as ongoing part-time or near-full-time contract work tied to a specific project's lifespan. Be honest in the screen about your weekly hours and time zone; underdelivering on stated availability is the most common way these engagements end early.

What they look for

Project or operations management experience with distributed contributors
Advanced Google Sheets / Excel — pivots, lookups, tracker design
Quality review and feedback delivery against written rubrics
Clear asynchronous written communication
Analytical comfort with throughput and accuracy metrics
Familiarity with LLM data annotation or evaluation workflows (preferred)

Before you apply

Would you pass this screen?

Drawn from what this listing actually requires. Answered in your browser — nothing is sent to us, and we do not store your answers.

Do you have at least two years of professional experience in project management, operations, or a role where you owned tracking, quality metrics, or team coordination?
Project management or operations experience · preferred
Are you proficient in Google Sheets, including formulas, conditional formatting, and basic data organization?
Fluency in Google Sheets · required
Have you managed, coordinated with, or led contributors, freelancers, or teams working remotely or asynchronously?
Remote or distributed team experience · preferred
Are you available to work during core US business hours (9 AM–5 PM EST, Monday–Friday) without scheduled conflicts?
No scheduled conflicts with US hours · required
Have you worked in a role where you regularly analyzed data, built reports, or made decisions based on metrics or trends?
Analytical mindset · preferred

0 of 5 answered

Included with this listing

Interview prep — 12 questions

Matched to what this platform's screening actually measures. Answer each out loud in 60–90 seconds: claim → concrete example → implication.

What it tests: Whether the candidate has hands-on experience building or maintaining performance dashboards (likely in sheets or similar), not just delegating that work, and whether they can spot anomalies in data before they become crises.
A strong answer: They name a specific tool stack (Google Sheets, Airtable, etc.), describe how they structured the metrics themselves, and give a concrete example of a metric trend that signaled a problem early—not a vague 'I monitored quality'.
Where people lose it: Saying they 'used Sheets' without explaining what they actually built in it, or describing quality tracking as purely subjective review without data—they'll lack the analytical infrastructure this role demands.