AI agent for recruiters
Interview Panel Composition Check Agent
Deliver an approved interview panel for each loop that has no conflicts, trained interviewers and a balanced mix.
What it does
Building an interview panel is a scramble. The recruiter needs the right people, but some have a conflict with the candidate, some have not finished interviewer training, and the panel may lean too heavily toward one team or background. For each interview loop, the agent reads the proposed panel and the candidate file, checks each interviewer for conflicts, training status and recent interview load, and measures balance across function, level and background. When a check fails it suggests a swap from a pool of trained interviewers. A swap can break the schedule, so the agent re-checks calendars, reschedules, and runs every check again. It stops when the panel passes. The recruiter always approves the final panel and any invite. Edge case: a swapped-in interviewer is free but already ran four loops this week, so the agent drops that option and tries the next name.
How it works
Follow the arrows from top to bottom. The orange dashed arrow is the loop: when a check fails, the agent goes back and tries again.
Read the steps as a list
- Interview loop opened for a candidate
- Read panel, candidate file and competencies to cover
- Check each interviewer for conflicts, training and weekly load
- Score the panel for balance across function, level and background
- Does the panel pass every check?If not: Pick a trained, conflict free swap from the pool that still covers the competencies. Back to step 3.
- Check calendars and reschedule the loop for the swap
- Does the new schedule work and does the changed panel still pass?If not: Drop that swap and try the next candidate on the list. Back to step 3.
- Recruiter approves the final panelThe agent waits here for your OK.
- Send invites after approval and log the panel
- Approved panel and scheduled loop
How it decides
It passes a panel only when every interviewer is trained, conflict free, within the weekly load cap and the mix meets the balance rule. Otherwise it picks the closest swap that keeps the competencies covered.
- A conflict (shared employer, relative, current reporting line) removes the person with no exception
- An interviewer without completed training cannot be on a panel unless the recruiter overrides in writing
- No interviewer takes more than 4 loops in a week
- Swap candidates are ranked by competency coverage first, then by availability
Make it yours
Every agent is a starting point. You choose these settings for your own situation.
- Weekly interview load cap (default 4)
- Balance rule: functions and levels required per panel
- Training that counts as complete
- Number of swap options to try (default 3)
What keeps you in control
It always asks you first
- Recruiter approves the final panel
- Recruiter approves invites before they go out
Hard limits
- Never contacts a candidate or interviewer without recruiter approval
- Never uses protected characteristics to pick a swap, only the stated balance rule
It stops when
- Done: panel passes every check and is approved
- Stop: no trained, conflict free swap exists, so the recruiter is told which gap remains
Set it up
We guide you through the set-up, step by step
Members get the full set-up guide for this agent. No technical skills needed: you copy, paste and upload.
- One set of instructions to paste into your AI, with the clicks for ChatGPT, Claude, Microsoft 365 Copilot, Gemini and Grok
- The agent then walks you through connecting your own data, one source at a time
- A downloadable copy with the flow chart, the rules and the full guide
