AI Interviews

AI Interviews is Skima AI's structured interview solution that lets recruiters evaluate candidates through AI-led conversations before moving them to the next stage of hiring. Recruiters can use AI Interviews to assess communication, technical knowledge, problem solving, and role fit in a consistent and scalable way.

Each interview is recorded, transcribed, and scored automatically. Recruiters can review candidates faster and make better hiring decisions backed by clear evidence rather than impressions from manual calls.

Navigate to 'AI Interviews' from the left sidebar to access the main dashboard.

Dashboard

The AI Interviews dashboard gives a full overview of all interview activity at a glance. It shows what is scheduled, what is pending, and what is ready to review, without switching between tools.

The four summary cards at the top show:

• Total Interviews across all jobs,
• Pending interviews that have been sent but not yet completed,
• Completed interviews ready for review, and
• Latest Interview showing the most recent activity for quick access.

Info

The page is divided into two tabs: Upcoming Interviews for sessions that are scheduled but not yet completed, and Past Interviews for sessions that have already been conducted.

Filters

Both tabs share the same filter set. Use Candidate Name or Email to find a specific candidate, Job Name to narrow interviews to a specific role, Status to filter by interview stage, and Start Date and End Date to view interviews within a specific time range.

Schedule an AI Interview

AI Interviews are created from inside a job's candidate list, not from the AI Interviews dashboard directly.

1. Navigate to 'Jobs' from the left sidebar and open the job for the candidate you want to interview.
2. In the candidate list, click the three-dots menu on the candidate's row and select 'AI Interview'.
3. The Create AI Interview modal will open for that candidate.

Note

You can also schedule interviews from the View Candidates tab inside a job. Refer to the Tabs and Filters guide for the full candidate actions available inside a job workspace.

Create AI Interview Modal

The modal shows the candidate name, email, and the job they are being considered for at the top so you can confirm the correct record before proceeding.

1. Select an existing Interview Flow from the dropdown if one has already been created for this role or round. Reusing an existing flow keeps evaluation consistent across candidates for the same position.
2. If no suitable flow exists, click the '+' icon to create a new one. See Create a New Interview Flow below.
3. Set the Event Date and Time for when the interview should be sent or made available to the candidate.
4. Click 'Create AI Interview' to schedule it.

Warning Always confirm the scheduled date and time before creating the interview. Incorrect scheduling can cause candidate confusion and missed sessions.

After the interview is created, a confirmation modal appears with the candidate's interview link, login email, and password. Copy and share these directly with the candidate. Always verify all three details before sending to avoid access issues.

Create a New Interview Flow

If no suitable flow exists, the flow creation process has three steps.

Step 1: Interview Duration

Select the interview length in hours and minutes. The duration shapes the number and depth of questions the AI generates. Click 'Next' to continue.

Step 2: Evaluation Criteria

Configure the interview structure and scoring areas.

i) Select the Round Number to indicate where this interview sits in the hiring process. Choose the Interview Type such as Technical, Behavioural, General, or Case Study based on the role requirements.

ii) Select the Interview Mode that controls how the session behaves:

• Standardized uses the same set of questions for every candidate, keeping evaluation consistent.
• Adaptive adjusts the flow based on the candidate's responses during the session. If a candidate gives a strong answer, the AI asks deeper follow-up questions automatically.
• Conversational runs the interview for the full set duration and lets the AI manage the pace dynamically, going as deep as time allows.

iii) Choose the Evaluation Criteria the AI should use to generate questions and score the candidate. Core criteria are always included. AI-generated criteria are suggested by the system based on the JD and interview type.

iv) Use Select All to include all available criteria at once.

Warning

Do not skip reviewing evaluation criteria. The selected criteria directly control the questions generated and the scoring output. Incorrect criteria will lead to poor interview quality and weak candidate assessment.

v) Click 'Generate Questions and Next' to proceed.

Step 3: Interview Details

Review and complete the final setup.

i) Enter a Title for the interview, typically the job title or a role-specific name. The Evaluation Criteria and Interview Info from the previous steps carry forward automatically.

ii) Review the Questions generated by Skima AI. Edit any questions that need to be more role-specific and add additional questions if needed.

Tip

Always review AI-generated questions before saving the flow. Small edits can make the interview more targeted and improve the quality of candidate evaluation significantly.

iii) Select the Voice for the AI interviewer and the Language for the session. If your preferred language is not available, contact [email protected].

iv) Use Additional Info to provide company context the AI should use when a candidate asks about the organization during the interview.

v) Use Additional Instructions for specific guidance on how the interview should be conducted, such as focus areas or expected answer depth.

vi) Add a Redirect URL if the candidate should be sent to a specific page after the interview ends.

vii) Write an Intro message the candidate sees before the interview begins and an Outro message that appears after the session is complete.

ix) Click 'Create AI Interview Flow' to save the flow. It will now be available in the Interview Flow dropdown when scheduling future interviews for the same role.

Upcoming Interviews

The Upcoming Interviews tab shows all sessions that are scheduled but not yet complete. Use it to monitor active candidates, confirm interview activity is progressing, and spot any sessions that may need follow-up.

Tip Check Upcoming Interviews regularly to catch delays early and reduce candidate drop-offs before incomplete sessions accumulate.

Past Interviews and Candidate Report

Once a candidate completes their interview, they move from Upcoming to Past Interviews. This is where recruiters open the full report and review the AI-generated evaluation.

Click on a candidate's name whose status shows Processed to open their detailed interview report.

Interview Status Reference

Status	What It Means
Scheduled	Interview created and assigned. Session has not started yet.
In Progress	Candidate is currently taking the interview.
Completed	Candidate has finished the session. AI analysis is queued.
Processed	AI analysis is complete. Scores, transcript, and video are ready for review.
Incomplete	Candidate started but did not finish. Review to decide whether to reschedule.
Failed	Could not be completed due to a technical or connection issue.
Rescheduled	Original time was changed. New session has been planned.
Cancelled	Interview was removed before completion. Record remains for reference.
No Show	Candidate was expected but never joined the session.

Candidate Interview Report

The report combines the AI score, transcript, video recording, and evaluation breakdown in one view.

i) The header shows the Candidate Name, Email, Job Title, Status Badge, and Overall Score so recruiters can quickly assess the result before reading the full detail.

ii) Transcript Panel shows the full conversation with time-stamped blocks and speaker labels. Use it to find specific moments without watching the entire video.

iii) Video Panel shows the recorded session. Hover over the audio visualization bars below the video to jump directly to a specific answer in text format without scrubbing through the recording.

iv) Scorecard Tab gives a structured breakdown across all evaluation areas used in the interview flow, including Communication, Skills and Experience, Problem Solving Methodology, Technical Implementation Skills, Domain Conceptual Knowledge, Adaptability and Resilience, Teamwork and Collaboration, Enthusiasm and Motivation, and Scale and Complexity.

Additional role-specific criteria may also appear depending on the flow configuration.

v) Questions Tab shows each AI question and the candidate's response separately, useful for answer-by-answer review without reading the full transcript.

vi) Proctoring Tab flags unusual session behavior such as screen changes or unexpected activity that may affect the reliability of the interview.

Warning Do not rely only on the scorecard when making a hiring decision. Always review the transcript, questions tab, and proctoring details before moving a candidate forward or rejecting them.

Tip Start with the overall score for a fast first impression, then use the scorecard and transcript together for deeper evaluation. When sharing results with a hiring manager, use the transcript and scorecard together so the discussion stays evidence-based.