What the work involves

You will be given real estate documents and AI-generated material about them, and asked to judge whether the analysis holds up. In practice that means reading a lease abstract and checking whether the escalation clause was summarized correctly, opening a pro forma in Excel and verifying that the exit cap and debt service assumptions are internally consistent, or reviewing a broker-style investment memo and flagging where the comparable set is inappropriate. Output is written: a structured rubric score plus a short justification that another reviewer could follow without your context.

Some tasks run the other direction — you produce the reference material. That might mean drafting a realistic property management scenario, building a small underwriting model, or writing the correct answer to a question a model got wrong. Both directions reward the same instinct: knowing which errors are cosmetic and which would actually cost a client money or blow a closing.

What the platform screens for

  • Specificity over titles. Mercor's screen is AI-led and follows up. Saying you "worked in commercial real estate" invites three follow-up questions; naming asset classes, deal sizes, and the documents you personally produced does not.
  • Document fluency. Whether you can tell an estoppel certificate from an SNDA, read a rent roll critically, or spot a mis-stated NOI without a calculator.
  • Rubric discipline. Whether you can score against written criteria rather than against your own preferences, and whether you can articulate the difference.
  • Written clarity. Justifications are read by people outside your specialty. Dense jargon with no explanation scores poorly.

Logistics

Fully remote and asynchronous, open to candidates based in the US, Europe, UK, or Canada. Work arrives in project batches — some weeks are dense, some are quiet — and most contributors treat it as part-time alongside existing practice, typically committing 10–20 hours weekly. There is no fixed schedule, but projects carry deadlines and throughput expectations, and quality audits are routine. Pay bands are observed on this platform and vary by project, specialization, and task complexity; they are not a guarantee.