AI-led and recruiter-led interviews for lean hiring teams · Private beta
Recruiter AI helps lean hiring teams run AI-led async interviews, recruiter-led live calls, and role-specific technical or domain screens - then turns each conversation into a scorecard tied to exact evidence, so you can evaluate more candidates without adding recruiting headcount.
●5 free interview credits · no card · human-reviewed recommendations
3
INTERVIEW MODES
0
AI AUTO-REJECTS
100%
SCORES CITE EVIDENCE
NOT A 99-CENT ONE-WAY SCREEN
Use async video if all you need is a recording. Use Recruiter AI when your team needs role-specific probing, live follow-up, recruiter-led conversations, and reviewable evidence.
INTERVIEW MODES
The entry wedge is replacing messy first-round screens. The workflow also supports technical, domain-specific, and recruiter-led interviews where humans need better evidence without more admin.
Candidates complete a structured voice/video interview from a secure link, with no account required. The AI asks follow-ups from your locked guide.
Your recruiter or hiring manager leads the call. The platform hosts the room, can record with consent, and shows the AI copilot as a visible participant. Transcript material is created or imported after capture.
Use structured questions for debugging, systems thinking, GTM judgment, product sense, support scenarios, or domain-specific reasoning.
HOW IT WORKS
The candidate experience is conversational. The recruiter experience is calm: every recommendation comes with the evidence needed to challenge it, whether the interview was AI-led or recruiter-led.
Say what a strong hire needs to prove, what risks to probe, and what topics to avoid. The system turns that into a structured interview guide.
YOU → AI
“”
Send an AI-led async link, schedule a recruiter-led live room, or use the same rubric for a technical or domain-specific screen.
You get a scorecard with citations, confidence bands, gaps, integrity context, and system-fault notes. The recommendation helps; the recruiter decides.
EVIDENCE-LINKED SCORECARD
SEE IT IN ACTION · AI-LED OR RECRUITER-LED REVIEW
No magic. Every score is anchored to what was said, what was missing, and what the system could or could not trust.
CANDIDATE · RECORDED SESSION
AUDIO · CANDIDATE
INTERVIEW MODE · HUMAN-LED WITH AI COPILOT
Recruiter leads · AI is visibly present
01:55 · AI
Walk me through an outage you owned end‑to‑end.
04:19 · MAYA
Last year our write path stalled across two regions. I was on call.
07:11 · MAYA
First thing: lowered the blast radius - rolled back the one region.
10:04 · AI
What were you watching for while you bisected?
12:57 · MAYA
Tail latency on the replication queue. Fanout first, then primaries.
16:47 · AI
What would you do differently?
19:40 · MAYA
Start with the obvious cache. We over‑built the read layer.
SCORES · EVIDENCE-LINKED
INTEGRITY
WHAT THE SCORECARD LOOKS LIKE
SENIOR BACKEND ENGINEER · TECHNICAL + DOMAIN SCREEN
SCORES · 3 AREAS
← linked to transcript · 04:12
← linked to transcript · 08:40
← linked to transcript · 16:02
TRANSCRIPT · EVIDENCE
04:12 · CANDIDATE
“I didn't wait for someone to tell me it was mine - I took the pager and wrote the postmortem the next morning.”
08:40 · CANDIDATE
“First thing I did was lower the blast radius - rolled back the one region, then started bisecting from there.”
16:02 · CANDIDATE
“We over‑built the read layer. If I did it again I'd start with the obvious cache and prove we need more.”
INTEGRITY CONTEXT · SHARED WITH CANDIDATE
Strong ownership and debugging signal. Advance to deeper system design.
AI NEVER AUTO‑REJECTS
The platform recommends. You decide.
Integrity signals are visible to the recruiter and the candidate - never used to silently drop anyone.
EVIDENCE‑LINKED SCORING
No mystery numbers.
Each rubric dimension references the interview material that earned it. Transcript, notes, gaps, and context stay inspectable.
NO DEEPFAKE AVATAR
The AI is clearly an AI.
Candidates see their own webcam and a branded audio visual - no synthetic face pretending to be human.
PRODUCT
01
Bring applicants in from bulk PDFs, CSVs, manual adds, or public apply links, then skim descriptive cliffnotes before you invite.
02
Candidates answer in a real voice conversation on their own time. The AI asks follow-ups when an answer is thin, vague, or promising.
03
Send a secure private link. Candidates attest identity, consent to capture, pass device checks, and start without creating an account.
04
Run human-led calls with video, screen share, explicit capture consent, and an optional AI Copilot that is visible to everyone.
05
Probe debugging, trade-offs, product judgment, support scenarios, sales discovery, or domain reasoning with a structured rubric.
06
Click any score to jump to the exact moment that earned it, with a confidence band and the gaps the evidence does not cover.
07
Change weights, pin must-haves, add dealbreakers, save proven guides as templates, and finalize before anyone opens the room.
08
Move candidates to shortlist, hold, reject, or follow-up with reviewer notes and a decision trail your team can audit later.
09
Focus changes, off-screen time, paste events, and answer-quality signals are review context, never automatic rejection rules.
10
Dropped connections, provider lag, and reconnects are logged as system faults and left out of scoring. They are never counted against the candidate.
11
Candidates see their own camera and a branded audio visual. The AI is clearly an AI, with no synthetic human face.
12
Download scorecards, transcripts, citations, and integrity signals for your hiring packet. Native ATS sync comes after launch.
FROM THE BETA
“We use it for async screens and recruiter-led technical calls.
The review finally shows what actually happened in the interview.”
Priya Ravindran
HEAD OF TALENT · SERIES B CLIMATE CO.
0
interview modes for AI-led, recruiter-led, and role-specific screens
Start with the mode that matches the role
0
candidate rejections made automatically by the AI
Recommendations are always recruiter-reviewed
0%
of rubric scores are tied to cited interview evidence
No unsupported scorecard numbers
PRICING
One credit is one completed interview — AI-led, recruiter-led, or technical — up to 45 minutes, with transcript, evidence-linked scorecard, and recruiter review. You’re only charged when there’s usable interview material: never for a no-show, a cancellation, or a system-fault-only session.
PILOT
5 INTERVIEW CREDITS
Run one real hiring loop, end to end.
Start free →TEAM
40 INTERVIEW CREDITS / MONTH
For lean teams interviewing every week.
Choose Team →SCALE
150+ INTERVIEW CREDITS
For talent orgs and high-volume roles.
Talk to us →1 CREDIT = ONE COMPLETED INTERVIEW UP TO 45 MIN · DEEP INTERVIEWS OVER 45 MIN USE 2 CREDITS · USD · CANCEL ANY TIME · DATA IS YOURS TO EXPORT
FAQ
No. The first wedge is replacing messy first-round screens, but the same workflow supports technical, domain-specific, and recruiter-led interviews. It is not a full onsite, coding-test, or final-decision replacement.
Candidates usually dislike silent recording boxes. AI-led interviews are live voice conversations with a clearly AI interviewer, no fake avatar, and a human recruiter making every decision. Recruiter-led rooms feel like a normal interview, with the platform handling capture and evidence.
No. Candidates join from a signed private link, attest that they are the invited candidate, consent to capture, run a browser device check, and start the interview without registering.
Yes. Recruiter-led rooms let a recruiter or hiring manager host the call, share screen, record with consent, and show the AI copilot as a visible participant. A transcript can be created from a recording or imported after the call; evidence-linked review starts only when that material exists.
Use bulk PDF drag-drop, CSV import, manual add, or public apply links. The platform creates descriptive cliffnotes for each application so recruiters can decide who to invite without turning resumes into black-box scores.
Stored encrypted in your workspace for as long as you want them. Candidates can request deletion at any time and you'll be notified. We never train on customer data. Period.
Not at launch. You can export scorecards, transcripts, citations, and integrity signals for your hiring packet. Greenhouse, Ashby, and Lever are the first native integrations planned after launch.
No. Ever. The platform recommends; you decide. Integrity signals are shown to both you and the candidate, never used to silently drop anyone.
That's on us, not the candidate. Every provider slowdown, drop, reconnect, or audio failure is logged as a system fault, shown on the report, and excluded from scoring. It can never count against the candidate. The candidate even sees which minutes were affected.
Candidates attest they are the invited person and will complete the interview themselves, over a signed private link. We surface identity and integrity confidence signals for a human to weigh, but we do not do biometric ID, ID uploads, face or voice matching, and we never claim a "verified identity." That stays a human judgment backed by evidence, never an automated gate.
Those tools are useful when you only need recordings at scale. Recruiter AI is for teams that need structured probing, live follow-up, recruiter-led calls, role-specific rubrics, evidence-linked scoring, and a recruiter-owned decision.
Because the cost is not just file storage. Each completed interview can include realtime AI, hosted rooms, audio/video capture, transcription, stability logging, screen-share evidence, and post-interview scoring. The value is recruiter time saved and better decision evidence.
Type II in progress (audit Q3 2026). DPA and DPIA support available today. Multi‑region and EU data residency are on the post‑launch roadmap, not available yet.
PRIVATE BETA
Start with one role. Brief the system. Invite candidates to AI-led screens or schedule recruiter-led calls. Review scorecards where every recommendation is tied to interview evidence.