"Domain expert" is a term every major AI training platform uses but rarely defines clearly. It's also the bucket where the rate jump from generalist is largest — often 2x or more. Here's what it actually means in 2026 and how to qualify.
The platform definition
Across Outlier, Mercor, Surge AI, and Turing, "domain expert" generally requires three things:
- A verifiable credential in a specific field — license, degree, or industry-equivalent experience.
- Active or recent practice in that field. Currency matters more than tenure; a doctor who hasn't practiced in 12 years pays less than one currently working.
- Subspecialty depth. "Doctor" qualifies; "pediatric oncologist with 15 years of clinical trial experience" qualifies as senior specialty.
The domains that pay most
- Medicine. MDs, RNs, NPs, PAs — verified through licensing boards. $90–$160/hr.
- Law. JDs with active or recent practice. Verified through bar admissions. $80–$140/hr.
- Quantitative finance. CFAs, quant fund alumni, PhDs in financial math. $100–$200/hr.
- Pharmaceuticals / biology. PhDs and industry researchers. $80–$130/hr.
- Tax / accounting. CPAs, EAs with active practice. $70–$110/hr.
- Mental health. Licensed therapists, psychologists. $70–$110/hr.
- Education. Teachers with subject-matter depth. $50–$80/hr.
Verification mechanics
Each platform verifies differently:
- Mercor: Most thorough — license number lookup, sometimes contacts your most recent employer.
- Surge AI: License number + sample of recent work product.
- Outlier: License number + a verification call for medical/legal specialties.
Verification adds 1–3 weeks to onboarding. Plan for the delay.
What "industry-equivalent experience" means
You don't always need a formal credential. Some domains accept industry depth:
- Quant finance: 5+ years at a buy-side fund, hedge fund, or proprietary trading firm typically substitutes for a CFA.
- Tax: 5+ years at a Big Four firm with verifiable role substitutes for a CPA.
- Medical research: 5+ years at a pharma/biotech with named published work substitutes for a PhD.
- Legal: 5+ years as a paralegal or compliance officer is usually not sufficient; legal domain specifically wants JDs.
Common positioning mistakes
- Burying the credential in education. Lead with it in your summary.
- Listing "physician" generally without subspecialty. Subspecialty depth raises rates 20–40%.
- Not noting current practice. "Currently practicing — 30+ patient encounters per week" matters; "MD, 2014" doesn't tell the screener if you're current.
- Applying for general coding tracks first. If you have a domain credential, apply for the specialty track directly.
The domains where supply is shortest
Three domains where qualified contractors are extremely scarce in 2026, leading to top-of-band rates:
- Quant finance with active recent buy-side experience.
- Pediatric / emergency medicine MDs willing to do contractor work.
- JDs with intellectual property, patent, or healthcare regulatory specialty.
Bottom line
"Domain expert" in AI training means a verifiable credential plus active practice plus subspecialty depth. If you have any of those, lead your profile with the specialty — generalist applications under-deliver compared to what specialty tracks pay. See full domain expert pay across the major platforms.