Contributor quality, from recruitment to removal.
Small on purpose: the same senior team that scopes your pilot is the one running it, and every contributor working on your data is accountable to the same standard.
Vetted in, monitored throughout, retrained or removed when needed.
This runs alongside our broader workforce model — this is the day-to-day mechanics of keeping quality consistent.
Recruitment
Candidates are sourced against the demographic, language, or skill profile your program needs — not pulled from a generic crowd pool.
Eligibility checks
Every candidate is screened against defined eligibility criteria before being considered for a program.
Verification, where appropriate
Identity or background verification is applied when the program's data or subject matter calls for it.
Training on your guidelines
Contributors are trained on your specific instructions and edge cases, not a generic task template reused across clients.
Qualification tests
Contributors have to pass a qualification test on your task type before they're allowed to touch live data.
Golden tasks
Known-answer tasks are mixed into live work to continuously check accuracy, without contributors knowing which is which.
Continuous monitoring
Performance is tracked for the life of the engagement, not just checked once during onboarding.
Performance scoring
Contributors are scored against defined quality metrics, and that score determines whether they keep working on your program.
Retraining
Contributors who fall below the bar are retrained and re-tested before being allowed back on live tasks.
Removal from program
Contributors who don't improve after retraining are removed from the program — quality isn't negotiated after the fact.
This is the same mechanism behind our five-stage QA system — see About for how it fits into the broader quality story.
Have a data requirement?
Give us the specification — we'll build the operation around it.
Start a Pilot →