AI agent for career coachs
Mock Interview Practice Loop Agent
Core interview questions answered at or above the bar, with approved best answers saved
What it does
Practicing interview answers without scoring them lets weak answers stay weak. This agent picks questions for your target role, reads or listens to each answer, and scores it against a clear rubric: relevance, evidence, structure and length. If an answer falls below the bar, it asks the same topic again from a new angle and scores the new try. It keeps track of what passed. A session ends when every core question clears the bar or time runs out. The best version of each answer is saved for you to approve as final. Edge case: you keep passing behavioral questions but missing technical ones, so the agent shifts the next round toward technical questions.
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
- Practice session begins
- Build a question list from the role and posting
- Ask the next question
- Score the answer against the rubric with notes
- Does the answer meet the pass mark?If not: explain the gap and ask the same topic from a new angle. Back to step 3.
- Record the best version of each passed answer
- Have all core questions passed or has time run out?If not: pick the weakest remaining topic and continue. Back to step 3.
- Job seeker approves the final answer versionsThe agent waits here for your OK.
- Saved answer bank and a list of weak topics
How it decides
Each answer is scored on a rubric with a pass mark. Weak topics are repeated from a new angle until they pass or time ends.
- Pass an answer when it scores at or above the bar on every rubric line
- Repeat a failed topic up to three times
- Shift the next round toward the weakest question type
- End the session when time is up even if some topics remain
Make it yours
Every agent is a starting point. You choose these settings for your own situation.
- Pass mark (default: 3 of 4)
- Session length (default: 30 minutes)
- Question mix (default: behavioral and technical)
- Maximum retries per topic (default: 3)
What keeps you in control
It always asks you first
- Job seeker approves which answer versions are saved as final
Hard limits
- Never presents scores as predictions of hiring results
- Never makes up facts for the person's answers
It stops when
- Done: all core questions pass and answers are saved
- Stop: time limit reached, with weak topics listed for next time
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
