Prepare
- Open the workbench: go to the web workbench; it loads once the system verifies a valid subscription.
- Environment: Chrome with microphone permission; headphones and a quiet room are recommended.
- Costs: creating your profile is free; the custom question bank costs 10 credits (charged once per JD); the voice interview costs 200 credits, with the dual-track report included.
Step 1: Create your job-seeker profile (free)
On the Profile page, paste or fill in your resume information. Question generation, interview scoring, and job decisions are all benchmarked to this profile — the more complete it is, the more targeted your practice. The workbench sidebar keeps a "Job-seeker profile not created" notice until you finish it.
Step 2: Generate the custom question bank (10 credits)
On the Mock interview page, paste the JD for your target job and click Generate question bank. You get a bank of your own with 5 dimensions and 10+ questions — self-introduction, project deep-dives, technical checks, behavioral questions, and your questions for them — each with what it probes and follow-up directions.
Read it through in writing first and work out the skeleton of each answer before going on mic. The same JD is charged only once — reviewing it repeatedly is not billed again.
Step 3: Voice practice (200 credits, report included)
- Confirm the cost, then start: the page clearly shows the 200 credits; billing starts only after you click start and pass the microphone check.
- Answer like it is the real thing: the AI interviewer converses in real time with a neural voice, and the first question usually starts from a project you have worked on. Mention an architecture choice and it asks "why did you pick that at the time?" — the gap between finishing a sentence and hearing the next one is about 1.3 seconds, close to the rhythm of a real conversation.
- About 20 minutes, 12–15 exchanges: one fully structured conversation. Being nervous is normal — that is exactly what practice is for.
- The report is generated automatically at the end: no extra steps, no extra charge.
Step 4: Read the dual-track report
Content track: what you said
Per-question scores with actionable advice. Every suggestion points to a specific question number and your exact words — "you spent 3 minutes on question 3 without getting past the background" is far more useful than "delivery could improve." Vague answers come with "a better answer" examples.
Delivery track: how you said it
Computed from word-level speech-recognition timestamps: net speaking rate, filler-word frequency, pause distribution, repeated restarts. Measured values, not model impressions — recomputing the same interview gives identical results. Score anchors are public, and what each score means is spelled out in the report.
After improving against the report, run another session on the same JD (the question bank is reused free — you only pay the interview's 200 credits) and watch the scores move.
FAQ
Does the 200 credits for a mock interview include the report?
Yes, the report is included at no extra charge; billing starts only once your microphone is ready.
Can I start an interview without a profile?
No. The question bank and the scoring are benchmarked to your profile — create a free profile on the Profile page first.
How long is the interview recording kept?
Recordings are deleted automatically after 30 days; the delivery metrics are measured and computed when the report is generated.
What if it misrecognizes what I said?
The report is based on speech-recognition transcripts, so individual words can be wrong; the scoring rules require that delivery not be penalized for transcription noise, and if you disagree you can practice another session.