The Voice Call Mode enables your company to conduct structured interviews with candidates through an AI-powered virtual interviewer. The conversation takes place in real time, via audio, capturing communication nuances and reasoning that resumes and written tests cannot reveal.
Why use the voice call?
Identifying what sets a candidate apart requires dialogue, not just answers. With this mode, you observe confidence, clarity of ideas, and the ability to think on their feet before the first in-person interview, while still scaling the process without compromising quality.
What sets the voice interview apart:
Active listening: the AI emits brief verbal cues to create a natural conversational environment.
Automatic voice detection (VAD): the system identifies when the candidate starts and stops speaking, with no buttons required.
Smart interruption: if the candidate wants to add something while the AI is speaking, the system immediately prioritizes the human voice.
Scalability: multiple candidates can be evaluated simultaneously using the same criteria.
How the interview works for the candidate
The experience is conducted entirely by the AI in three stages:
Introduction: the AI introduces itself, provides context on the topics, and creates a welcoming environment so the candidate can show their best.
Conversation and follow-up: for very brief answers, the AI gently requests more details or concrete examples, ensuring a fairer evaluation.
Closing: once all topics are covered or the configured time is reached, the AI wraps up cordially and ends the call.
During the call, the candidate has access to:
A real-time audio level indicator to confirm they are being captured.
A processing status ("Listening" / "Processing") to know when they can speak.
A mute button for quick pauses without ending the interview.
Example: Customer Success interview
Here is how the AI conducts a session in practice:
The AI asks about account retention strategies.
The candidate responds and the AI signals active listening with brief verbal cues.
If the answer is vague, the AI probes further: "Can you give a real example of a metric you improved?"
After closing the topic, the AI transitions smoothly to the next one or thanks the candidate and ends the call.
Technical requirements
To ensure a smooth experience, advise candidates to use:
Browser: updated Chrome, Edge, or Firefox.
Headphones: required to avoid echo and ensure the AI only analyzes the candidate's voice.
Stable connection: a reliable internet connection to prevent delays in the conversation.
How to create an assessment with voice call
Go to the Question Library in the main menu.
Click New question and select the type AI Conversational Interview.
Under Conversation format, select Voice (audio call).
Choose the interviewer Avatar. An audio preview is available before confirming.
Add at least one Topic with a name and guiding question (required). Use evaluation hints and follow-up questions to guide the AI in probing shallow answers.
Set up the Evaluation criteria in the review guidelines section (see below).
If needed, add a Script with context instructions for the interviewer.
Publish the question and add it to the assessment through the Sections tab.
The maximum number of conversational interviews per assessment varies by plan. Once the limit is reached, the system displays a warning and prevents adding new interviews.
Evaluation criteria
Criteria are internal guidelines configured in the question that are not shown to the candidate. They define what the reviewer should observe in each response and serve as the basis for the AI's automatic review after the interview.
Each criterion has a name, a description of what to evaluate, and a weight from 1 to 5 indicating its relevance to the final score. Essential criteria receive a high weight; complementary aspects receive a lower weight. A criterion can be marked as extra to reward exceptional answers without penalizing candidates who do not reach that level.
Well-defined criteria make the review more objective, reduce variation between reviewers, and increase the reliability of results. The platform offers automatic criteria generation based on the configured topics, which can be used as a starting point and adjusted according to the needs of the process.
Tips to get more from the results
Ask for real stories: configure guiding questions to seek concrete examples. High-performing candidates stand out when describing situations they have experienced.
Use for soft skills: verbal clarity, persuasiveness, and emotional intelligence emerge naturally in conversation.
Calibrate topics: choose themes that challenge systemic thinking and assess the candidate's strategic vision through speech.
How to review the interview results
After the candidate completes the interview, the response becomes available on the platform with the status Awaiting review.
Go to Assessments and open the participant's result.
Locate the conversational interview response in the side panel.
Review the conversation transcript and listen to the audio recording.
Check the Automatic review to see the analysis generated by the AI, if this feature is available.
Click the edit icon next to Score to assign a grade from 0 to 100 and add a memo with your observations.
The final score is defined manually by the recruiter. The automatic review serves as a support tool, not a decision-maker.
Automatic review
At the end of each interview, the AI analyzes the conversation transcript and evaluates the candidate's performance based on the criteria configured in the question. The result appears in the review panel and includes an overall summary, a text-based feedback, and the individual score for each criterion with its respective weight.
The automatic review is available for both conversational interview modes — chat and voice call. The use of this feature, as well as the conduct of conversational interviews themselves, may incur additional credit costs depending on interview volume. Please consult your plan or contact support for more details.