Enterprise FAQ.

Yes — field programs are scoped to the cities, regions, or demographics your project specifies. Coverage is confirmed per engagement rather than fixed in advance, so tell us the locations you need and we'll scope against them.

Yes. Recruitment criteria — age range, location, language, device type, and other attributes relevant to your model — are defined as part of scoping, before a program starts.

Programs typically start as a small pilot, then expand in stages once quality and turnaround are validated against your spec. Talk to us about your specific volume and timeline and we'll tell you what a realistic ramp looks like.

Language and dialect coverage is confirmed per engagement based on your requirement and available collector or annotator coverage. Tell us the languages your program needs and we'll scope against them directly.

Yes — annotator training and QA calibration are built around your guidelines, not a generic taxonomy we apply to every client.

Yes, custom taxonomies are standard. We adapt training materials, task design, and QA review to match your schema.

Through defined PII handling and redaction practices, role-based access controls, and consent management appropriate to your project. See Security & Trust for the full list of data-handling practices.

Automated duplicate detection and similarity checks run as part of our standard QA pipeline, alongside human review — not as an afterthought at final delivery.

Through a five-stage system: contributor validation, automated checks, human QA, calibration, and client acceptance. See the Quality section on our homepage for the full breakdown.

Yes — human review sits at multiple stages of our QA pipeline, not only at final sign-off.

Yes, NDAs and confidentiality workflows are standard for enterprise engagements.

Yes — every engagement can start as a small pilot scoped against your spec, so you can evaluate quality and turnaround before committing to a larger volume.

Delivery is coordinated around whatever secure storage or transfer method your program requires. Tell us your preferred destination when we scope the pilot and we'll confirm compatibility directly.

Reporting cadence — daily, weekly, or milestone-based — is set during scoping to match how your team tracks delivery.

Rejected data goes through a rework and root-cause process rather than being resubmitted as-is. This feedback loop is built into the client acceptance stage of our QA system.

Requirements are expected to evolve. Scope changes are handled through the same direct line you have to the team running your program — not a change-request queue.