A live AI copilot for your human interviewers
Interview Copilot passively transcribes your live video interviews, tracks which competencies have been explored, and suggests follow-up questions in real time. Your interviewer still runs the conversation — Kaaira just makes sure no competency falls through the cracks.
Already a customer? Sign in to your dashboard
What Interview Copilot does during your live interview
Passive transcription of both sides
Interview Copilot transcribes the interviewer and the candidate in real time, with speaker attribution and timestamps. The conversation continues as normal — Copilot listens, it does not speak. Your candidate experience is unchanged.
Live competency coverage tracker
Each role competency gets a coverage state — NOT EXPLORED, LIGHTLY EXPLORED, or WELL EXPLORED — based on the topics the interviewer and candidate have actually discussed. You always know which competencies still need probing before the call ends.
Follow-up question suggestions on tap
When you want a probe, Kaaira generates a follow-up question grounded in what the candidate has just said and weighted toward the under-explored competencies. The suggestions are aware of your job context, your competency framework, and the conversation so far — not generic interview clichés.
Recorded video interviews are not the answer
One-way recorded video interviews remove the conversation. Live AI screens remove the human. Interview Copilot keeps the human interviewer and adds the AI scaffolding — coverage, probing, and post-call synthesis — that makes the interview rigorous.
Connected mode — linked to the job and candidate
Pick the job profile and the candidate application. Copilot pulls the competency framework, job context, and candidate name automatically. Sessions land on the candidate's profile alongside their assessment scores so reviewers see the complete picture.
Ad-hoc mode — paste a JD, get started
For interviews outside the formal pipeline — reference checks, calibration calls, exploratory chats — paste a job description and start. No job profile required, no candidate record needed. Useful for hiring managers who want the structure without the setup.
Coverage report at the end of every call
Post-call, every competency gets a 0–100 coverage score and a question count. You see, at a glance, where the conversation went deep and where it skimmed — so the next round (or the rejection note) is grounded in evidence, not impression.
Capability insights, not just a transcript dump
Kaaira synthesises the conversation into structured strengths, concerns, and signals — drawn from what the candidate actually said. The interviewer reviews the synthesis instead of re-reading 40 minutes of transcript before debriefing the hiring team.
Email the summary to the hiring manager
One click sends the post-call summary, capability insights, and coverage report to the hiring manager (or the candidate's owner in the ATS). The debrief happens with shared evidence in everyone's inbox.
Augments human judgement, never replaces it
Copilot does not score the candidate, does not make a hire/no-hire call, and does not recommend a decision. Its job is to make the human interviewer more rigorous. The hiring decision stays with the human, with better evidence to support it.
From setup to scored session in three steps
Open Copilot in a browser tab next to your video call. That's the whole integration.
Pick connected or ad-hoc
Connected: choose a job profile (and optionally a specific application) — competencies, job context and candidate name are pulled automatically. Ad-hoc: paste a job description and name the candidate. Either way, set-up takes under 30 seconds.
Run your interview as normal
Open Copilot in a browser tab next to your video call. It listens passively. As the conversation unfolds, the competency coverage panel updates live; when you want a follow-up question, tap Suggest and Kaaira returns a probe grounded in what was just said.
Wrap up with a synthesised debrief
Close the session and Kaaira produces a written summary, capability insights (strengths, concerns, signals), and a per-competency coverage report. Email the package to the hiring manager, or keep it on the candidate's profile alongside their assessment scores.
Augment the interviewer, don't replace them
Interview Copilot is a structural support, not an authority. Every design choice keeps the human firmly in the loop.
Copilot listens, candidates speak to humans
The AI does not speak to the candidate. The candidate experience is a normal human-led conversation. Copilot is in a tab on the interviewer's screen — invisible to the person being interviewed.
No hire/no-hire recommendation
Copilot does not score the candidate and does not output a recommendation. The strengths, concerns, and signals are observations grounded in transcript evidence — not a verdict. The hiring manager decides.
Suggestions are competency-aware, not generic
The follow-up question prompt is built from your job context, your competency framework, the current coverage state, and the most recent turns of the actual conversation. Kaaira will not suggest a generic 'tell me about a time…' question when the competency is already well explored.
Tenant-isolated sessions
Each session is bound to an organisation, with RLS-enforced row-level isolation. Copilot sessions are visible to the creator and to org owners and managers — never across organisations, never to other interviewers without permission.
Same scoring engine as voice interviews when used together
If you also run a Kaairo AI Voice Interview for the same role, Copilot's coverage report uses the same competency framework. Hiring managers see human-led interview signal and AI-led interview signal against the same rubric.
Explore Related Solutions
Discover more ways Kaairo can help your organization hire and develop talent.
Frequently Asked Questions
Is Interview Copilot the same as HireVue or one-way recorded interviews?
No. Recorded interview platforms remove the conversation — candidates record themselves answering pre-set questions, the AI analyses the recording. Copilot keeps the live, two-way human conversation. It runs alongside your normal video call, transcribes both sides, and helps the interviewer probe more thoroughly. The candidate is talking to a human, not to a recording app.
Does the candidate know AI is being used?
That is your call. Copilot is invisible to the candidate — it runs in a browser tab on the interviewer's side and does not speak. Most teams disclose its use the same way they would disclose recording or note-taking, but the technical setup does not require disclosure.
What does 'competency coverage' actually measure?
Each role competency starts as NOT EXPLORED. As the interviewer and candidate discuss topics, mentions are attributed to competencies; once a competency has been mentioned a handful of times, it moves to LIGHTLY EXPLORED, then WELL EXPLORED. Post-call, each competency gets a 0–100 coverage score and a question count so you can see exactly where the conversation went deep.
What is the difference between connected mode and ad-hoc mode?
Connected mode is for interviews inside the pipeline — pick a job profile (and optionally a candidate application), and Copilot pulls the competency framework, job context, and candidate name automatically. Ad-hoc mode is for interviews outside the pipeline — calibration calls, reference checks, exploratory chats. Paste a job description, name the candidate, and start. Same Copilot, no setup overhead.
Can I see the AI's follow-up suggestions before asking them?
Yes. Suggestions appear as text on the interviewer's screen. The interviewer chooses whether to ask the suggested question, rephrase it, or ignore it. Nothing is asked of the candidate without the interviewer's choice.
How is the post-call summary generated?
The full transcript, job context, competency framework, and candidate name are passed to GPT-4o, which produces a structured summary plus capability insights (strengths, concerns, signals) drawn directly from the candidate's words. The prompt explicitly forbids fabrication or hire/no-hire recommendations.
Which plan is required?
Interview Copilot is available on the Enterprise and Enterprise+ plans. Organisations on Starter or Professional can run AI Voice Interviews (asynchronous, Kaaira-led), case studies, SJTs, and MCQs — but the live human-led copilot is a Enterprise-tier feature.
Who in my organisation can use it?
Owners and managers. The creator of each Copilot session has full access to their own sessions; org owners and managers can view all sessions inside the organisation. Members cannot create or view Copilot sessions.
See Interview Copilot run a live interview
Request a demo and we will run a live interview with Copilot listening — and walk you through the coverage report and post-call summary at the end.
Talk to us about your team
Share a few details about your org and what you're trying to solve. We'll follow up with a short call or an async walkthrough of how Kaairo can plug into your hiring, benchmarking, or L&D workflows.