What the work actually involves
You author scenarios that a competent practitioner could resolve but a language model will likely botch: a valuation where the obvious comps are contaminated by a distressed sale, a purchase agreement whose contingency timelines interact badly, a zoning question turning on a nonconforming use, a deal structure where seller financing changes the tax picture. Each item needs a defensible correct answer, a written rationale, and usually a set of plausible-but-wrong alternatives that reflect real mistakes agents and clients make. The second half of the job is evaluation — reading model output and explaining precisely where the reasoning fails, not just marking it wrong. "Cites a habitability standard that doesn't exist in this state" is useful feedback; "answer is inaccurate" is not.
What the platform screens for
AfterQuery's screen is AI-led and follow-up heavy. Expect to name your licensing state and license status, describe deal volume and property types you've actually closed or managed, and then get pushed one or two layers deeper on whatever you claim. Generalists who say "residential and commercial" and then can't discuss a cap rate reconciliation or an estoppel certificate tend to stall there. The screen also probes jurisdictional honesty — real estate law is state-specific, and writers who present local practice as universal create bad training data. Designations (CRS, GRI, CCIM, CPM) and any proptech or publication history help, but they do not substitute for transaction depth.
Logistics
- Fully remote, asynchronous — you pick your hours, work is claimed from a queue
- Hourly, observed at $50–100/hr depending on specialization and calibration performance; not guaranteed
- No fixed weekly minimum in the posting, but throughput and consistency drive continued assignment
- Expect an unpaid or lightly paid calibration task before volume work opens up
- Contractor engagement, project-based; volume fluctuates with client demand