Terms.
The short version: we bill for runs that produce a validated finding, we never bill for a refusal, and nothing acts on your customers without your approval.
Last updated 2026-07-25
The service
Hunter-Seeker ranks rows in a table by likelihood of a yes/no outcome, with calibrated scores, out-of-time validation, and a per-run audit record.
Metering
Usage is metered in two units: runs and entities scored. A run that ends in an honest-empty refusal costs zero credits.
Human approval
Actions drafted by a standing agent are staged in an approval outbox. Nothing is delivered to a third party without an explicit approval recorded in the ledger.
What we do not claim
- Drivers are labeled as model associations. We do not claim causation.
- We do not claim to be the most accurate model available.
- We do not offer autonomy beyond the approval model described above.
Availability
Audit SLAs are available on the Governed tier. Other tiers are provided on a commercially reasonable efforts basis.
Contact
Questions about these terms: sales@hunter-seeker.net