Understand the output before choosing an import route
The Oracle-oriented export is CSV with PersonNumber, StartTime, StopTime, MeasureValue, PayrollTimeType, LocationCode and Comment. Attendify supplies the worker’s payroll ID as PersonNumber, captured endpoints as timestamps, calculated paid hours as MeasureValue and the site name as LocationCode.
This is not an HCM Data Loader package or a universally accepted Oracle time-event payload. Your integration team must map the source into the supported receiver interface for its environment.
Validate values as well as column names
A familiar column label does not make a value valid. Check the person number, configured time type and location reference. A site display name may need conversion to a receiver code. Start/stop timestamps are absolute instants; verify how the target handles local dates and overnight work.
Configured break deductions can make the exported hours differ from elapsed endpoint duration. Decide how the receiver uses those values so it does not duplicate a break or recalculate a conflicting total.
Separate source review from Oracle approval
Attendify applies its daily-hours calculation and excludes blocking attendance days from policy-based exports. Its correction approval is a source-data workflow. It does not substitute for Oracle’s configured time-card approval and payroll transfer requirements.
Oracle’s payroll time approval documentation explains receiver-side approval tasks. Review the relevant configuration with your Oracle owner rather than treating a downloaded file as transfer-ready payroll.
Run a small acceptance set
Include a mapped worker, missing payroll ID, overnight shift, configured break and approved correction. Reconcile source counts/hours, transformed rows, receiver rejections and accepted results. Preserve the receiving run identifier for later investigation.
Attendify does not currently promise scheduled delivery or a certified Oracle connector. The handoff is complete only when your receiving process has validated and accounted for the data.