For Creators

Practices Overview

Learn how Aural Practices turns an existing interview into a repeatable coaching workspace with voice practice, AI feedback, suggested answers, and score tracking.


What Practices Does

Practices lets you rehearse against the questions in an existing interview without creating real candidate sessions. Each run opens a focused coaching experience where the AI asks the interview questions, listens to your answer, scores the attempt, and gives specific feedback before you move on.

Practice session with AI coaching feedback, score, and suggested answer panel
Practice session with AI coaching feedback, score, and suggested answer panel

Interview-native practice

Practice uses the same interview questions your team already wrote, so rehearsals stay aligned with the actual role, rubric, and question order.

Voice-first coaching

Focused practice starts in voice mode, records the answer, builds a transcript, and evaluates both answer quality and delivery signals.

Feedback after every answer

Each submitted answer produces a score, verdict, strengths, improvements, missing signals, and next-step coaching.

Suggested answer support

The side panel can generate a role-aware sample answer when practice context is available.

Where to Find Practices

  • Global Practices page - open /practices to review practice runs across the interviews in your current project.
  • Interview prep tab - open an interview, then go to the Practices tab to manage context and review runs for that specific interview.
  • Focused practice mode - click Practice interview to open a dedicated practice session in a new tab.

Practice is not a candidate session

Practice sessions are saved separately from real interview sessions. They do not consume session time, appear in candidate results, or replace completed candidate interviews.

What Gets Saved

Aural saves practice sessions and answer attempts so you can measure progress over time. Saved data includes the interview, practice mode, status, start and completion timestamps, total duration, submitted answers, per-answer feedback, attempt count, average score, and best score.