Technical Interview for Project Managers? Rehearse It with AI Step by Step

You are three minutes into a video call. The interviewer leans in and asks why your last project slipped by six weeks when the plan looked fine at kickoff. You start explaining, but your answer drifts into generalities about communication. The silence on the other end tells you it was not enough. That moment is exactly what you can rehearse beforehand, without a friend who happens to be a hiring manager.

What the AI actually does when you ask it to interview you

A conversational assistant like ChatGPT, Gemini, or Claude does not know your projects. It predicts likely wording based on patterns it learned from many texts, including job descriptions and interview guides. When you assign it the role of a senior interviewer and paste in your project details, it generates questions that match that role and follows up based on what you write back. It is not judging you. It is producing plausible pressure so you can practice responding to it.

The trick is to keep each turn short. If you ask for a full mock interview at once, you get a tidy list of questions and no follow-up. If you ask for one question at a time, the assistant has to react to your answer, which is closer to the real thing.

Test one: the polite request that gets you nowhere

The naive prompt

Act as a project manager interviewer. Ask me some questions about my experience.

This feels helpful, but it is a quiz sheet, not an interview. The questions are generic, there is no follow-up, and nothing forces you to defend a decision. In a real technical interview for a project manager role, the interviewer will interrupt, ask for numbers, and challenge your reasoning. This output trains you to recite, not to think on your feet.

Test two: the calibrated prompt that pushes back

The calibrated prompt

You are a senior technical interviewer hiring a project manager for a platform migration. Ask me exactly one question at a time. After each answer, challenge one specific claim I made and ask for a number, a trade-off, or a decision I would change. Do not praise me. Do not move on until I give a concrete answer. Start with the hardest question about my last project.

Now the assistant behaves like a hiring panel. It asks for one thing, refuses to accept a vague story, and demands a number. When you answer, it will pick apart the weakest part of your reply. That is the rehearsal you actually need, because it exposes the answers you have not fully thought through.

Master template to copy and reuse

Fill in the brackets with your own details. The more specific your context, the sharper the questions.

You are a senior technical interviewer for a [ROLE] position at a [COMPANY TYPE] company. My background: [PASTE 3-4 LINES ABOUT YOUR PROJECTS, TEAM SIZE, AND TOOLS]. Your job: ask me one question at a time about [TOPIC 1], [TOPIC 2], and [TOPIC 3]. After every answer, challenge one specific claim and ask for a number, a trade-off, or a decision I would change. Do not compliment me. Do not accept vague answers. If I dodge, repeat the question. Keep each reply under 80 words.

Quick steering commands when the AI gets too soft

• That answer was too easy. Ask me the same question again but harder, and do not let me off with a general statement.
• You accepted my answer without a number. Push back and ask me for the specific figure I used to make that call.
• Stop summarizing my answer back to me. Just ask the next question and keep the pressure on.

How you know the rehearsal worked

You know it worked when you can answer the same hard question three different ways without repeating yourself, and when you can name the number behind a decision without pausing. A good sign is that the AI's follow-up no longer surprises you. Then switch to the free voice mode in the mobile app and run the same simulation out loud. Speaking forces you to hear your own filler words and rushed endings, which is the part a typed chat cannot fix.