Skip to main content
AI voice interviews · 25+ languages

Interview every candidate.In their language.Impossible to fake.

Interloop runs AI voice interviews in 25+ languages, scores every answer against your rubric, and blocks AI-assisted cheating with continuous proctoring — face, gaze, liveness, voice. Free for your first 50 interviews.

No credit cardSetup in 5 minGDPR + DPDP compliant
interloop · live 14:23
REC
EN
English
Gemini Live
translating live
AI

Walk me through a decision you're proud of.

M

I killed a feature half the team loved — the churn data said it was the problem, not the fix.

scoring liverubric · 96/100
Problem-solving92
Communication88
Depth95
face verified gaze on-screen 0 flags
CLEAN
25+ spoken · 95+ understood
EnglishEspañolFrançaisDeutschItalianoPortuguêsNederlands日本語한국어中文العربيةTürkçePolskiРусскийहिंदीবাংলাதமிழ்తెలుగుമലയാളംಕನ್ನಡमराठीગુજરાતીਪੰਜਾਬੀไทยTiếng ViệtBahasaEnglishEspañolFrançaisDeutschItalianoPortuguêsNederlands日本語한국어中文العربيةTürkçePolskiРусскийहिंदीবাংলাதமிழ்తెలుగుമലയാളംಕನ್ನಡमराठीગુજરાતીਪੰਜਾਬੀไทยTiếng ViệtBahasa
0+
Voice languages
spoken · 95+ understood
0
Proctoring signals
face · gaze · liveness
0%
Proctoring pass rate
first attempt
0m
Avg time-to-shortlist
vs 5+ days traditional

Hear it yourself.

Start a 60-second live interview — pick your language, no signup.

Try a live interview

Built for teams hiring across India & the GCC — engineered, not a ChatGPT wrapper

Gemini LiveMediaPipeface-api.jsGDPR + DPDPFace data never leaves the browserContinuous proctoring

Hiring still runs on spreadsheets, gut feel, and guesswork.

Resumes lie.

A polished CV tells you nothing about how someone actually thinks under a real question.

ChatGPT cheats your screens.

Right now, candidates are using AI on the side of your interviews — and most tools can't see it.

Recruiters lose Saturdays.

Phone screens don't scale. Your best people burn weekends on calls that a rubric could score.

00Meet Nova

Not just an interviewer. An AI HR agent that shows its work.

Nova screens every applicant, runs the interviews, scores them against your rubric, and routes the close calls to you — then logs why, keeps you in charge of the hard decisions, and hands each candidate proof they own. One agent across the whole loop, not a stack of disconnected tools.

01Explainable

It shows its work

Every automated decision is logged with the score, the threshold, and plain-language reasoning — and rolls up into an auditor-ready governance report.

02Fair

You keep the hard calls

Require your sign-off before Nova declines a candidate or sends an offer. Candidates see their score breakdown and can request a human review.

03Verifiable

Proof of Talent

Every interview mints a public, verifiable credential — tier, score, and evidence — that stands on its own, independent of any one employer.

04Yours

Owned by the candidate

The proof belongs to the person, not the platform — portable across every application. Verification before trust.

And it spans the whole lifecycle — an org chart it can reason over, internal-mobility matches for open roles, and gap-focused learning paths — while syncing Greenhouse, Lever, Workday and 50+ more, so nothing lives in a silo.

01The hiring layer

One real interview, in any language, scored against your rubric.

Voice conversations beat forms. Gemini Live asks real follow-ups across 25+ languages, catches evasions, and scores every answer against your rubric — with a transcript excerpt behind every number. Candidates interview in the language they think in. You get evidence, not guesses.

Free Fly · AI-ledStructured · your question bankAssisted · live co-pilot
01

Voice AI in 25+ languages

Gemini Live · 25+ languages

Spanish, Japanese, Arabic, Mandarin, Hindi, Malayalam, Portuguese — and 20+ more. Candidates pick their native tongue; scoring stays in yours.

02

Evidence-based scoring

Rubric · Evidence · Per-dimension

Every number comes with a transcript excerpt. Every rubric dimension shows you where — and why — the score landed there.

03

Auto-pipeline engine

Rules · Automation · Fair

Set thresholds once. Interloop auto-invites above 80, shortlists above 90, gently rejects below 40. You review the interesting middle.

04

Comparison & decisions

Teamwork · Decisions · Audit

Side-by-side candidates with aligned rubric bars. Comments, @mentions, decision trails. Move forward in two clicks.

interloop.tech / pipeline
Live
Active role

Senior Software Engineer

Remote · Global·247 applicants·12 shortlisted
18
pass rate
Candidate
Status
Score
AD
Aria Delgado
Senior Engineer·ES
Offer sent
94
KT
Kenji Tanaka
Product Designer·JA
Interview done
91
CL
Camille Laurent
DevOps Lead·FR
Shortlisted
87
PN
Priya Nair
Full-stack Dev·ML
In review
82
FA
Fatima Al-Rashid
Security Analyst·AR
Awaiting
76
Live now
Aria · ES · 12:24
“…diseñé un sistema que redujo el tiempo de desarrollo un 40%…”
Problem-solving +12
Proctoring96 · CLEAN
Fraud signals0 / 7
Face verified
PProctoring

Your screens are being cheated right now. Ours can prove they aren't.

ChatGPT on the side. Someone else taking the call. A recorded video played back. Deepfakes. We stop all of it — six independent signals running continuously from first hello to final goodbye. Most platforms verify identity once. We verify every second.

01
Continuous face verification
MediaPipe landmarks vs the enrollment photo, every second.
02
Gaze tracking
468-point face mesh catches looking off-screen — flags the moment.
03
Liveness challenges
Random blink & head-turn tests prove a live human, not a replay.
04
7-signal AI fraud detection
Timing, fillers, vocabulary — reveals AI-assisted answers.
05
Browser monitoring
Tab switches, paste, fullscreen exits, DevTools — all logged.
06
Tamper-proof evidence
Every violation timestamped with severity + screenshot. Dispute-ready.
Score
96/100
Face
100%match
Gaze
94%on-screen
Events
0violations
Session · 18:42 elapsed
CLEAN
96
/100
Proctoring integrity
Face match
100
Gaze tracking
94
Liveness
100
Browser behavior
92
Event log
00:03face_verified
04:21liveness.passed
08:17face_verified
16:30browser.fullscreen_exit (2s)
16:32browser.fullscreen_restored
18:42session.end

Privacy-first by design. Face data never leaves the candidate's browser.

Start free — 50 interviews
02Proof of knowledge

Prove your team can still do the job — not just that they once could.

A completed course is a receipt, not a result. Upskill runs short AI voice assessments on your SOPs and products — so each person recalls and explains it out loud. You get proof of who knows what, where the gaps are, and whether last quarter's training still holds.

Before

100% complete. 0% proven.

Support
45
52
38
Engineering
68
55
72
Sales
40
35
58
Operations
52
48
44
— “Did you finish the course?” “Yes, technically.”
— Audit finds 40% gaps nobody reported.
— Managers guess who needs training.
After

Knowledge you can prove.

Support
88
92
85
Engineering
94
89
91
Sales
86
82
88
Operations
90
93
87
Every gap found, targeted, and re-verified next cycle.
Study guides generated from your own docs, in their language.
Compliance proof on demand.
01Measure

Recall, out loud.

A 10-minute AI voice assessment on your SOPs and policies. Saying it out loud — not clicking it — is what proves someone knows it.

02Pinpoint

Every gap, by name.

See exactly which person, team, and topic is weak. A live heatmap that finds gaps before an auditor does.

03Close

Training from your own docs.

Gemini reads your knowledge base and each person's gaps, then writes the study material to fix them — in their language.

04Verify

Proof it stuck.

Re-assess on a cadence and watch recall climb. You don't just close the gap — you prove it stayed closed.

03The process

Setup in 5 minutes. First interview in 10. Hire by Friday.

1
Step 01

Create role

Pick a rubric template or build yours. Choose voice language. Set proctoring level. Five fields, three minutes.

avg: 3 min
2
Step 02

Share apply link

Candidates apply via OTP + email. Auto-pipeline screens resumes and triggers invites by score.

handles 100s/day
3
Step 03

AI voice interview

Gemini Live conducts the conversation. Face + gaze monitored. Every answer scored in real time.

~22 min/session
4
Step 04

Review & hire

Open evidence-backed shortlists. Compare candidates. Send offers directly. Track acceptance.

days not weeks
Why teams switch

Every platform says they use AI. Most are wrappers.

01

Voice AI in every language.

25+ spoken, 95+ understood. Gemini Live native — candidates answer in the language they think in, not yours.

25+ languages · Gemini Live
02

Engineered, not assembled.

Not a ChatGPT wrapper. Gemini Live for voice, MediaPipe for gaze, face-api.js for verification — each picked for the job.

Gemini · MediaPipe · face-api
03

Actually stops AI cheating.

Six proctoring signals incl. a 7-signal AI-fraud sub-detector, continuous from first hello to final goodbye. Most platforms verify once.

Continuous · 6 signals
04

Interview to continuous learning.

Hire and Upskill share one AI, one rubric language, one dashboard. The platform grows with your team.

Hire + Upskill · One platform
04The commercials

Transparent pricing. No quote walls.

Every number on the page. Less than one recruiter's Saturday — per month, not per interview. A human screen runs ₹800–5,000 each; Interloop runs hundreds for a flat plan. Cancel anytime — your data exports as CSV.

Hiring for a role right now? Buy credits — no subscription.

Pay once · UPI / card · valid 12 months

No subscription — pay once, GST invoice included, no auto-debit, valid 12 months. Two ways to interview:

Express1 credit / 15 min

Fast, focused screening — ideal for high-volume first rounds.

Deep2 credits / 15 min

Full-conversation memory + deeper probing — for the candidates that matter.

Starter pack
100 credits
₹5,999+ GST
₹60 / credit
100 Express · 50 Deep
Growing team
250 credits
₹11,999+ GST
₹48 / credit
250 Express · 125 Deep
Serious hiring
500 credits
₹16,999+ GST
₹34 / credit
500 Express · 250 Deep
Agency / volume
1000 credits
₹29,999+ GST
₹30 / credit
1000 Express · 500 Deep

Same-Day Hire: turn it on per role and candidates go from apply → AI interview → offer letter automatically — no recruiter in the loop until the offer is out.

Need volume or enterprise terms? ↓
Enterprise
Custom

For large organizations — volume credits, custom terms, and dedicated support.

  • High-volume credit terms
  • Unlimited team members
  • SSO / SAML
  • Dedicated CSM
  • Custom integrations
  • SLA guarantee
  • White-label
Contact sales
05Questions, answered straight

Before you sign up.

We support real-time voice interviews in 25+ languages natively — Spanish, Japanese, Arabic, Mandarin, French, Hindi, Malayalam, and more — and the AI understands 95+ others. Our proctoring is continuous (face + gaze tracked throughout), not just at the start. And we offer transparent multi-currency pricing with no quote walls.
Every interview produces an integrity score from multiple signals — response timing, filler-word and vocabulary patterns, continuous face + gaze monitoring, tab-switch tracking, and random liveness challenges — and the result is a verifiable integrity certificate you can check. It raises the bar against AI-assisted answers and flags the ones that look off for human review.
Yes. Face verification runs entirely in the candidate's browser — their biometric data never leaves their device. Interview data is stored in regional clusters (Frankfurt, Singapore, Mumbai, or Virginia — your choice). GDPR + DPDP compliant with data export and deletion on request.
Our scoring is evidence-based — every score is backed by transcript excerpts. Tests show 85%+ correlation with human reviewer scores. AI catches things humans miss (consistency across questions, specific skill signals) and vice versa.
Voice interviews in 25+ languages via Gemini Live — including English, Spanish, French, German, Portuguese, Italian, Dutch, Japanese, Korean, Mandarin, Arabic, Turkish, Russian, Ukrainian, Polish, Romanian, Hindi, Bengali, Tamil, Telugu, Malayalam, Kannada, Marathi, Thai, Vietnamese, Indonesian, and more — plus conversational understanding of 95+ others. Candidates answer in their native language; you score in yours.
You don't need a subscription. If you're hiring for just a role or two, buy prepaid interview credits — pay once via UPI or card, no auto-debit, no mandate. Credits are valid for 12 months. There are two interview types: Express (1 credit per 15 minutes) for fast, high-volume screening, and Deep (2 credits) — our most intelligent interview, with full-conversation memory. Packs start at ₹5,999 for 100 credits, with bigger packs at lower per-credit rates. Teams hiring every month can still subscribe.
Both run as a real AI voice interview and produce a complete, rubric-scored report — the evaluation is just as rigorous either way. The difference is the live interviewer: Deep keeps the entire conversation in memory and probes deeper across everything the candidate said, ideal for senior or final-round candidates; Express runs a faster, focused screen that's perfect for high-volume first rounds. Express is 1 credit per 15 minutes, Deep is 2. You choose the tier per role when you set it up.
2-5 minutes. Sign up, create your first role (wizard-guided), set your rubric, and share the apply link. We don't require implementation consultants or enterprise onboarding calls.
Hire is for bringing new people into your team — AI interviews candidates, scores them, manages the pipeline. Upskill is for your existing team — instead of just delivering courses and tracking completion, it runs short AI voice assessments that prove what each person actually knows about your processes and products, pinpoints the gaps, auto-generates training to close them, and re-checks on a cadence so you can prove it stuck.
An LMS tracks completions — who clicked through which course. It can't tell you whether anyone retained it. Upskill measures recall: each employee explains your SOPs, products, and policies out loud in a short AI voice assessment, so you get proof of competence, not just a completion certificate. You see exactly where the gaps are, training is generated from your own knowledge base to close them, and everyone is re-assessed on a cadence to confirm the knowledge held.
Yes. Use Structured mode to define your own question banks with weighted rubrics. Or use Free Fly mode and let the AI conduct a dynamic conversation based on your rubric criteria.
At random intervals (5-10 minutes apart), the system might ask the candidate to 'blink twice' or 'turn your head slightly left.' This verifies a live human is on camera — not a pre-recorded video or deepfake. Takes 3-5 seconds.
Yes. We connect to your ATS through a unified Merge integration — push interview scores and hired candidates back to Greenhouse, Lever, Workday and 50+ others. Plus email, CSV import, WhatsApp & LinkedIn share links, a REST API, and webhooks. Coming: Slack and Microsoft Teams.

Have another question? Ask us — we reply within 4 hours

§ Newsletter

The hiring intelligence brief, weekly.

What's actually working in Indian + GCC hiring right now — a 3-min read every Tuesday. No spam, one-click unsubscribe.

Start free today

Fifty interviews on us. No card.

Setup takes five minutes. First interview in ten. Hire your first great candidate by Friday.

No credit card50 free interviewsSetup in 5 minCancel anytimeCSV export