Most people still use AI for interview prep the old way: "give me some interview questions" or "help me write my self-introduction." There's a better move — flip it around and let the AI play the interviewer: asking you questions one by one, pressing for details, then reviewing your performance. The prompt below does exactly that. Paste it into the chat box, fill in four variables, and your mock interview starts immediately.
Who this prompt is for
- You're preparing for experienced-hire, campus, or career-switch interviews and want to get used to the rhythm of being pressed with follow-ups
- You want to practice deep project dives (one project questioned three layers deep), behavioral questions (STAR structure), or interviews in English
- Not for people who just want to memorize a question bank: this trains expression and on-the-spot thinking, not recitation
The prompt: copy this block
You are an interviewer with 10 years of experience hiring for [job title], preparing for an [interview round] at a [company type].
Follow these rules:
1. Ask only one question at a time. Wait for my answer before asking the next one — never list all questions at once.
2. Start with "Please give a 1-minute self-introduction," then proceed in this order: project deep-dives → business understanding → behavioral questions, for a total of [number of questions] questions.
3. After each question, keep pressing with follow-ups based on my answer until the details are fully explored (e.g., "Why is this metric at that number?" "What would you do differently if you did it again?").
4. Do not give hints or evaluations while I'm answering. After all questions are done, give me a full review.
Format the review like this:
- An overall score (1–10) with a one-line verdict
- My 3 biggest point deductions, each illustrated with a quote from my actual answers
- For each deduction, a rewritten model answer
Let's begin: please let me start with my 1-minute self-introduction.Variables: swap in your situation
| Variable | What to fill in | Examples |
|---|---|---|
| [job title] | Your target role | Backend engineer, product manager, marketing operations |
| [company type] | Target company type — sets the questioning style | Big tech, startup, multinational |
| [interview round] | Which round you're practicing | Technical/business round, HR round, final round |
| [number of questions] | Questions per session, 5–8 recommended | 6 |
Tuning tips: getting the most out of it
- Paste your resume first: send your resume right after the prompt, and the interviewer's follow-ups will stick to your experience instead of drifting into generic questions.
- Set difficulty in one line: add "press me at one level above big-tech bar" for a harder session, or "keep it at new-grad campus recruiting level, no management questions beyond my scope" for entry-level prep.
- For English interviews: add "ask everything in English, I'll answer in English, and point out my English expression issues in the review."
- Stop it from "cheating": the most common failure is the AI dumping all questions at once. Rule 1 exists to prevent that; if it still does, just reply "one question at a time" and it will usually correct itself.
- Make the review worth reading: if the review is vague ("be clearer"), follow up with "for deduction #2, rewrite my original words into a perfect answer" — the gap becomes immediately visible.
Two common misconceptions
Q: Isn't it easier to just have the AI generate questions and memorize the answers?
A: Memorization works for quizzes, not interviews. Behavioral questions and project deep-dives test how you organize language and react on the spot — a memorized answer falls apart under follow-ups. Mock interviews train exactly that: staying coherent while being pressed, which no question bank covers.
Q: Can I trust the score the AI gives?
A: Don't take the number too seriously — it has no real baseline to compare against. What's genuinely valuable are the deductions and the model answers in the review: revise through two rounds and your self-introduction and project stories will get noticeably stronger. To see what human interviewers actually look for, compare with this recruiter's-side hiring prompt (/article/145): knowing both how to answer and what the other side is watching.