Interview is a separate area from the ticket queue, reached from its own tab in the sidebar. It runs mock interviews for an entry-level help desk role against an AI hiring manager, and it is intentionally self-contained: it does not use the same AI persona, prompts, or logic as the ticket-side chat and call features, because interviewing you is a different job than roleplaying a broken computer.
Sitting a mock interview
- Open Interview and choose a focus — Mixed, or a specific area like Troubleshooting, Customer Service, Security & Identity, Escalation, Documentation, or IT Fundamentals.
- Choose whether to answer by typing or by speaking out loud.
- Start the interview. The interviewer asks one question at a time and reacts briefly to what you said before asking the next.
- Answer honestly and in your own words — the interviewer never coaches you or reveals how you're scoring mid-interview.
- Click End interview when you're done, or once the interviewer has asked its planned set of questions and wraps up naturally.
When you end the interview, it is analyzed as a whole — a separate grading pass from the interviewer itself, so the same conversation gets both a live interviewer and an independent after-the-fact judge, the way a real interview loop might have separate people conducting and later reviewing it. You get an overall score, a summary, a list of what you clearly demonstrated, a list of what to work on, and a question-by-question breakdown that includes the answer the interviewer was hoping to hear for each one.
Every finished interview updates a skill profile tied to your account — which topics you're strong in, which you're weak in, and your running average score. The next interview you sit reads that profile before it asks a single question: it will spend less time re-testing something you've already shown you know, and weight new questions toward the areas you're weakest in. This is what makes the feature "customized to your skill gap and current level" rather than a fixed question bank — the questions genuinely change based on your own history.
The Practice screen, reached from the interview setup page once you have history, lists every individual question you've been asked before. You can reattempt any of them on their own, get it graded independently of a full interview session, and see whether your answer has actually improved — useful for drilling one specific weak question without sitting a whole new interview around it.
In voice mode, the interviewer — referred to in the interface as Riley Chen, IT Support Manager — speaks its questions aloud using the same text-to-speech engine the rest of the platform uses for AI end users, and listens to your spoken answer the same way calls do. In text mode, the same interviewer asks the same kinds of questions and applies the same grading; only the input and output medium changes, not the substance of the interview or how it's judged.
Interview is deliberately separate from every ticket-side AI feature, down to the prompt and persona level. The chat and call features roleplay a simulated employee with a broken computer; Interview roleplays a hiring manager evaluating you. They share the same underlying AI infrastructure because both need a live, responsive conversational voice, but nothing about your ticket-side communication history feeds into your interview skill profile, and nothing about your interview history changes how a ticket-side employee talks to you. It is a genuinely standalone practice space, low-pressure enough to attempt honestly, because a rough interview here has no bearing on your ticket queue or your career points.
You can choose a topic focus before starting — Troubleshooting, Customer Service, Security & Identity, Escalation, Documentation, or IT Fundamentals — or leave it Mixed and let the interviewer draw from all of them. Choosing a focus is useful when you already know which area you want to drill; leaving it Mixed is closer to what an actual interview loop looks like, since a real interviewer rarely stays confined to one topic for the whole conversation.