The Handshake AI Fellowship is structurally different from most of the platforms we track. Where Mercor and micro1 are spot-market placement engines, Handshake's Fellowship is a structured part-time engagement — defined hours, defined scope, a fixed term — built around academic credentials. It's run through Handshake, the university career platform, and explicitly targets graduate students, postdocs, recent PhDs, and early-career researchers.
Here's how the Fellowship works in 2026, what it pays, and who it fits.
What Handshake AI hires for
Fellowship roles center on domain-expert evaluation and reference generation across academic fields:
- STEM evaluation — math, physics, chemistry, biology, CS reasoning at graduate level.
- Research-grade reference answers — producing the gold-standard responses models train against.
- Specialist domain work — economics, statistics, and other research-adjacent fields.
Pay typically runs $75–$125/hr, with the cluster around $90–$100. Engagements are usually 10–20 hrs/week over a fixed term — which fits an academic schedule far better than spiky spot-market work.
The application process
Applications go through joinhandshake.com (the existing career platform) or directly via Fellowship opportunity pages. You submit a résumé and, for most roles, a short research statement or cover-letter equivalent. Interviews are often human-conducted (a research lead on a 20–45 minute call), though some Fellowships use AI-led screens. Note: many roles require U.S. work authorization — check the specific posting.
Who Handshake AI is best for
- PhD candidates and postdocs — the Fellowship is designed for you, and the predictable hours fit research schedules.
- Master's students and early-career researchers — a strong first platform before you have the senior credentials the higher-paying spot-market platforms expect.
- Anyone who wants predictable income over spiky high-rate placements.
Where Handshake AI fits vs. the rest
Handshake trades a lower ceiling than Mercor for predictability and an academic-friendly structure — the full head-to-head is in our Mercor vs Handshake AI comparison. PhDs should read our guide for PhDs, and the featured-platforms roundup covers the four-way picture.
Bottom line
If you're a graduate student, postdoc, or recent PhD who wants structured, predictable AI-training work at $75–$125/hr, the Handshake AI Fellowship is the best-fit platform in the market. Apply via Handshake and browse current listings on our Handshake AI listings page.