Quality and review
Rubrics, calibration, and what happens when work fails review.
Quality is measured against published rubrics, gold-standard items, and reviewer consensus — not self-assessment.
How work is scored
Programmes may include calibration tasks, inter-annotator agreement checks, and spot audits. Alignment with the rubric drives re-invites. Drift or rushed work triggers review or removal from the active programme.
Failed calibration
You may get extra training, a reduced scope, or removal from that cohort. Failing calibration is not an automatic ban from Harbor. Repeated failure in the same domain lowers match priority.
Removal from the network
Harbor may remove people for confidentiality breaches, fraud, repeated quality failures after feedback, harassment, or attempts to bypass payout or data-handling rules. Profiles may be paused while a dispute is investigated.
Repeat invitations
Deliver on time, follow the rubric, ask early, and keep calibration scores. People who handle complex eval work well are first in line when a client expands.
Need a programme scoped? Talk to Harbor.