AI recruiting coordinator automation handles approved-stage logistics: finding interviewer availability, offering timezone-safe slots, sending confirmations, managing reschedules, and chasing structured feedback. It should not rank candidates, infer suitability, reject applicants, or make employment decisions. The safest design minimizes candidate data, uses calendar permissions narrowly, supports accommodations, and gives recruiters a complete communication history.

Outcome and Non-Goals

The outcome is a consistent interview process with less scheduling delay and fewer manual handoffs. Recruiters should know the candidate’s stage, approved interview plan, participant roles, available windows, confirmation state, accommodation request, feedback status, and next owner.

The workflow does not:

• Decide who advances to an interview.

• Score a resume, interview, voice, face, personality, or protected characteristic.

• Reject or deprioritize a candidate.

• Replace a recruiter’s accommodation process.

• Reveal private calendar details beyond availability.

• Send material changes to the hiring process without recruiter approval.

The U.S. EEOC warns that algorithmic employment tools can unlawfully screen out people with disabilities and emphasizes reasonable accommodations. Keeping this workflow focused on coordination rather than evaluation materially reduces, but does not eliminate, employment and privacy risk (EEOC guidance announcement).

Inputs and Systems

Use only the data needed for coordination:

• Applicant tracking system stage and approved interview plan.

• Candidate contact details, timezone, language, and stated availability.

• An accommodation-request route handled by authorized staff.

• Interviewer roles, calendars, working hours, and backup panels.

• Approved interview duration, sequence, and location or meeting platform.

• Calendar free/busy data rather than full event details where possible.

• Message templates, sender identity, reminders, and cancellation policy.

• Feedback forms and completion deadlines.

• A communication and change log.

Google Calendar’s free/busy endpoint can return occupied time ranges without requiring event content. Its authentication guidance recommends choosing the narrowest scopes possible, which is the right permission model for scheduling assistance (Google Calendar free/busy API, authorization scopes).

Numbered Workflow

1. Receive an approved scheduling request. Require ATS stage, recruiter owner, interview type, duration, panel roles, candidate timezone, and target window.

2. Validate the interview plan. Check that required participants, sequence, buffers, location, and feedback form exist. Route incomplete plans to the recruiter.

3. Read availability narrowly. Query free/busy and working hours for approved interviewers. Do not retrieve or expose private event titles.

4. Generate feasible options. Apply timezone, working-hour, buffer, panel, accessibility, location, and sequence constraints. Mark any relaxation of a constraint.

5. Offer choices to the candidate. Use approved language, display timezone explicitly, provide an accommodation contact, and avoid first-come pressure unless the process requires it.

6. Confirm and create events. Recheck availability, create the approved event, add meeting details, notify participants, and update the ATS.

7. Handle rescheduling. Preserve the reason only when necessary, release canceled holds, generate new options, and alert the recruiter after repeated changes.

8. Send reminders. Use approved timing and allow candidates to correct timezone, location, or accessibility issues.

9. Track feedback completion. After the interview, send structured feedback tasks and reminders. Do not summarize feedback into an autonomous hiring recommendation.

10. Close the coordination task. Record attendance, feedback completion, unresolved issues, and the recruiter’s next action.