Career guide · Examples reviewed September 6, 2026
The same title can mean judging model responses, testing AI agents or helping a business use AI tools. Start with the responsibilities to find the work that fits you.
Review responses, compare answers against a rubric, explain errors and create training examples. Clear writing, careful judgment and subject knowledge matter. Some opportunities are freelance projects; others are employed roles.
Build workflows or agents and test how reliably they perform. Look for responsibilities such as evaluation suites, failure analysis and regression testing. These roles may require programming and experience shipping software.
Help teams choose tools, identify use cases, improve processes and learn new workflows. This is a broader career path. We include a role here when evaluating or improving model behavior is part of the actual job.
Meridial's AI Generalist freelance project describes reviewing model outputs and producing training data. It advertises an hourly rate and identifies contributors as independent contractors.
Columbia University's AI Generalist role instead focuses on consulting, request intake and helping university teams adopt AI. It is a full-time role with an annual salary range. It illustrates the broader use of the title; it is not counted in this board's matching openings.
These employer descriptions were checked on September 6, 2026. Availability and terms can change; the openings below are drawn separately from our current board.
4 current listings use “generalist” in the title, as of 2026-09-06. These roles already meet our evaluation, training or related-work criteria. This is a snapshot of our coverage, not the total market.
Seattle, WA$35–$100/hr
Remote (USA)Remote listed$45–$55/hr
Canada · Calgary +4Remote listed
Remote (United States)Remote listed$20/hr
Check whether the offer is employment, a fixed contract or access to project work. An advertised hourly rate does not establish how many paid hours will be available. Confirm location eligibility, assessment requirements, benefits and how work is allocated before accepting.
Use AI trainer hourly pay for contributor examples and the annual salary report for salaried roles. Neither is an estimate of what everyone with “AI generalist” in their title earns.
Our suggestion: choose one kind of work and build a small example of it. For human evaluation, compare model answers to ten questions using a written rubric and explain disagreements. For technical evaluation, build a repeatable test set and document failures. For AI adoption, demonstrate a workflow improvement with a measurable result.
You do not need to present yourself as an expert in every AI tool. Show evidence that you can do the tasks a particular employer needs. Read the actual qualifications: “generalist” does not automatically mean entry-level or no coding required.
Explore routes into AI evaluation → · Get the weekly hiring brief →