Skip to main content
Use this sequence to keep candidate work visible without turning Harly into a second inbox.
1

Start from the pipeline

Open the job pipeline and work from the oldest actionable applications. Use filters for owner, stage, department, or region instead of exporting candidate data to personal spreadsheets.
2

Review the candidate record

Read the application, resume, answers, notes, and activity timeline before changing the stage. Add a concise note that explains the decision and the next owner.
3

Assign the next action

Create a task for a missing review, use a scorecard for structured feedback, or move the application to the next stage. Harly updates open views through its realtime channel; a hard refresh is only a recovery path.
4

Schedule the interview

Create the interview with the attendees, timezone, duration, and video provider. Calendar and meeting integrations sync in the background. If a provider is unavailable, the interview remains saved and shows a warning that can be retried.
5

Communicate from the record

Send candidate email from the application or mailbox. Keep replies attached to the candidate thread and avoid forwarding sensitive resumes to personal accounts.
6

Close the loop

Send an offer, record the decision, or reject with the appropriate reason. For signed offers, verify the signing status from the document record, not from an unverified screenshot or email claim.

Collaboration rules that scale

  • Put the decision and next step in a note, not only in chat.
  • Use scorecards for comparable interview feedback.
  • Assign one accountable owner per application.
  • Treat AI output as a draft or signal; a human owns the hiring decision.
  • Export or delete candidate data only through an approved privacy workflow.
Configure an automation for routine notifications, but keep irreversible actions such as rejection, deletion, or offer withdrawal behind an explicit human step.