New
Government photo-ID face match
Candidates show a PAN, Aadhaar, passport or driving licence; the face match runs in their browser and must reach 70% before the interview starts.
For employers · white-label AI interviewer
InstaDataInterview The AI interviewer — your first round, fully automated.
Send one link. The AI verifies the candidate, holds an adaptive voice interview with a live coding round, probes answers that sound rehearsed, and hands your HR team a rubric-scored, explainable report — under your brand, from $99 a month.
- 8-role demoPlay the candidate — no signup.
- No card14-day trial, 5 interviews included.
- Your brandLogo, colours and support email.
9interview languages, from English and Hindi to Arabic and Japanese
8ready-made demo roles, technical and functional
30 mindefault interview — briefing to probing follow-ups
14-dayfree trial · no credit card · 5 interviews
Hands-free voice with barge-in
Defend-your-answer probing
Per-skill adaptive questioning
Python & JavaScript in the browser
Agentic coding mode
Photo-ID face match
Big Five behavioural profile
HR-approved interview blueprint
Bulk CV intake
Talent pool
Email & WhatsApp invites
Side-by-side compare
Explainable integrity index
Model card
Four-fifths adverse-impact check
Candidate appeals
Human override
White-label workspaces
Shipped in June 2026
Eight upgrades that make the first round harder to fake and easier to run.
Identity, behaviour and volume — the parts of screening that used to need a person in the loop — are now built into the interview itself. Each one is optional and set per job.
New
Big Five behavioural profile
An optional OCEAN, communication and work-style read, inferred from what candidates say and how they say it — no extra questions.
New
HR-approved interview blueprint
The AI proposes the plan; nobody can be invited until HR approves it. The approved plan sets difficulty, coding, language and voice.
New
Bulk CV intake
Drop in a stack of CVs (PDF, DOCX, TXT). Each is parsed and matched to the job with a score, strengths and gaps.
New
Talent pool
Reuse past candidates for new openings — one login can be a candidate for several jobs.
New
WhatsApp invites
Send credentials and the one-time interview link by email or WhatsApp, once your workspace connects them.
New
Candidate compare
Put finalists side by side — scores, skills and integrity — before you decide who moves on.
New
8-role public demo
Four technical and four functional roles anyone can try, plus an "AI answers for you" mode to watch a full run.
Inside the interview
It conducts the interview — it doesn't just record it.
One session covers the conversation and the technical screen. The AI listens, lets candidates interrupt, chases weak spots and checks integrity as it goes.
Hands-free, like a real call.
- Voice activity detection calibrates to the room's noise floor, so candidates just talk.
- Barge-in: interrupt the AI and it stops — confirmed within 380 ms.
- Interview in 9 languages; the interviewer voice and accent are set per link.
- Browser voice by default; self-hosted Piper, ElevenLabs or Azure neural voices can be switched on.
- Screen and camera are recorded together, with audio and a transcript, for HR to review.
"Defend your answer."
- Answers that look AI-generated or read out are always followed up — the candidate has to explain their own reasoning.
- A polished answer with a weak defence raises the AI-likelihood signal that feeds the integrity index.
- Per-skill ability estimates (a Rasch-style model) pick the next question: the least-known skill, at matched difficulty.
- Every answer is scored for relevance, depth, correctness and consistency against the job's rubric anchors.
The coding test, in the same session.
- Python and JavaScript run right in the candidate's browser — nothing to install.
- Agentic mode: AI tools are allowed, but the candidate writes a justification that is scored alongside the code.
- Paste bursts during the round are logged as integrity events.
- More languages can be run server-side on request.
Integrity you can explain.
- In-browser proctoring: tab switches, face missing, multiple faces, looking away, background noise and paste bursts.
- Object detection runs on the candidate's device; a second screen is flagged on Chromium browsers.
- An Integrity Index from 0 to 100 combines these events with each answer's AI-generated likelihood — and lists the top factors.
- Facial expressions are not used for proctoring or scoring.
A behavioural read, at no extra length.
- Big Five (OCEAN) traits plus communication and work-style, inferred from transcript content and speech delivery.
- No questionnaire and no extra questions — the interview stays the same length.
- Switched on per job; when off, the section simply doesn't appear in the report.
How it works
From job description to shortlist in five steps.
-
01
Upload the JD
AI parses title, skills and requirements and drafts a rubric of competencies with 1–5 anchors and weights.
-
02
Approve the blueprint
Review the AI-proposed interview plan. Candidates can only be added once HR approves it.
-
03
Invite in bulk
Upload CVs or pick from your talent pool. Credentials and a one-time link go out by email or WhatsApp.
-
04
The AI interviews
About 30 minutes: briefing, profile, skills, coding, follow-up and probing — with ID and integrity checks.
-
05
Review and decide
Open the report, compare finalists, override the recommendation if you disagree, export a PDF.
Human in the loop
The AI proposes. HR approves.
Before a single candidate is invited, the AI drafts an interview blueprint from the job and its rubric. Your team edits and approves it — and the approved plan drives difficulty, the coding round, the language and the voice.
01Profileexperience & ownership
02Skillsincident response · observability · Kubernetes
03CodingPython · medium · agentic mode
04Follow-up & probedefend-your-answer
Awaiting HR approval · invites locked
Identity
Know it's the right person.
Switch on a government photo-ID check for any job. The candidate shows a PAN, Aadhaar, passport or driving licence; the face match runs in their browser and must reach 70% before the interview begins. The match score sits at the top of the report.
For real interviews the document is kept with the record; in the public demo it is never kept.
The report
A recommendation you can defend line by line.
Every interview ends in a structured report: five core dimensions, per-skill results against your rubric, the integrity index with its reasons, an optional behavioural section and a recommendation from Strong hire to No hire. Export it as a PDF, or override it — with an audit trail.
78
Overall score · recommendation
HIRE
Core dimensions
Mental alertness80
Subject knowledge76
Skill level74
Communication82
Integrity84
Download PDFOverride recommendationCompare
Per-skill results · rubric anchors 1–5
| Incident response | 4 / 5 | |
| Observability | 4 / 5 | |
| Kubernetes | 3 / 5 | |
| Python (coding round) | 4 / 5 | |
| Communication | 4 / 5 |
Integrity index · top factors
Paste burst during coding−7
One answer read as AI-generated, defended well−5
Brief tab switch−4
Candidate compare
Finalists, side by side.
Line up the people who made it through — scores, skills, integrity and the AI's recommendation in one view — then let your team make the call.
| Candidate A | Candidate B | Candidate C | |
|---|---|---|---|
| Overall | 78 | 71 | 64 |
| Skill level | 74 | 72 | 58 |
| Coding | 4 / 5 | 3 / 5 | 3 / 5 |
| Integrity | 84 | 91 | 69 |
| Recommendation | HIRE | MAYBE | NO HIRE |
Try it yourself
Pick a role. Be the candidate.
The public demo has the same eight roles shown here — technical and functional. Take the interview yourself, or switch on "AI answers for you" and watch a full run. No signup, and the demo makes no hiring decision.
Technical
Functional
Four of the eight demo roles also switch on the photo-ID check and the behavioural profile.
Technical
sample question
Solution Architect
The AI might open with
Governance built in
Fair, explainable, and open to challenge.
An AI that screens people should be able to show its working. The compliance tools are part of the product, not a consulting project.
On facial analysis: no facial-emotion inference is used in scoring or proctoring. An optional expression signal for the behavioural profile is off by default, needs the candidate's consent and is blocked for EU candidates.
Rubric anchors
Each competency has calibrated 1–5 anchors and a weight, so scores mean the same thing for every candidate.
Model card
A transparency record of what the AI does, what it uses and its limits, ready for your compliance team.
Four-fifths check
Adverse-impact monitoring on outcomes, using the four-fifths rule.
Candidate appeals
Candidates can ask for human review; an admin resolves each appeal as upheld or overturned.
Human override
Reviewers can change any AI recommendation, scoped to their tenant and logged for audit.
Retention you set
Retention per workspace (365 days by default) and automatic purge of expired recordings.
White-label & multi-tenant
Your interviewer, your brand.
Every workspace carries its own name, tagline, colours, logo and support email — candidates see your company, not ours. Staffing firms, RPOs and job platforms can run many client workspaces from one console and resell the interviewer as their own.
- Admin, HR and candidate roles, with HR seats and password resets managed by your admin.
- A super-admin console for resellers to manage tenant workspaces.
- Bring your own AI key per workspace, plus your own email and voice settings.
- Shared, dedicated or on-your-server deployment — talk to us.
- Available on request: ATS sync (Greenhouse, Lever, Workday and more via Merge) and self-hosted speech-to-text.
Who it's for
Built for teams that screen at volume.
High-volume first rounds
Tech and functional roles alike — the demo covers sales, pharma, banking, marketing and compliance as well as engineering.
Staffing firms, RPOs & job platforms
A white-label interviewer you can put your own brand on and offer to every client.
Tech hiring teams
Merge the behavioural screen and the coding test into one session and one report.
Regulated employers
Rubric anchors, a model card, four-fifths checks and appeals — with dedicated-server options.
Which one do I need?
InstaDataInterview and InstaDataAssess share the same AI interview engine. The difference is who makes the final call.
You are here
InstaDataInterview
The AI interviewer — your first round, fully automated.
- Job to do
- One link, one AI interview, one report
- Who interviews
- The AI only
- Getting started
- Self-serve: free trial, plans from $99/month
Need panel rounds too?
InstaDataAssess
The interview platform — AI screens, your experts decide.
- Job to do
- Run the whole funnel: AI rounds, then live expert-panel rounds, then a governed decision
- Who interviews
- AI for volume; your experts on live video for decisive rounds
Pricing
Pay for the interviews you run. Start free.
Every plan includes adaptive voice interviews, structured rubric scoring and explainable integrity. 14-day free trial, no credit card.
Fits your volume
Free
$0 / month
5 interviews / month · 2 seats
- Adaptive voice interview
- Structured rubric scoring
- Standard email support
Fits your volume
Starter
$99 / month
50 interviews / month · 5 seats
- Everything in Free
- Coding round
- Integrity index & proctoring
- Multilingual interviews
Most popular Fits your volume
Growth
$299 / month
200 interviews / month · 15 seats
- Everything in Starter
- Agentic coding round
- Neural interviewer voice
- ATS integration (configurable)
- Priority support
Fits your volume
Scale
$799 / month
1,000 interviews / month · 50 seats
- Everything in Growth
- White-label branding
- Bias-audit & compliance pack
- Dedicated success manager
More than 1,000 interviews a month, a dedicated server or your own infrastructure? Talk to us. Prices in US dollars, per month.
Frequently asked questions
What is InstaDataInterview?
InstaDataInterview is an AI interviewer for employers. You send a candidate one link; the AI holds an adaptive voice interview with an optional coding round, checks integrity (and photo ID if you switch it on), and produces a rubric-scored report with a hire recommendation for your HR team to review. It runs under your own brand, from $99 a month, with a free plan and a 14-day trial.
Is it a mock-interview tool for job seekers?
No. InstaDataInterview is built for employers, staffing firms and job platforms who interview candidates. The only open-to-anyone part is the public demo at /experience, where you play the candidate in one of eight roles to see how it works. The demo makes no hiring decision.
How is it different from InstaDataAssess?
Both use the same AI interview engine. InstaDataInterview automates the first round: the AI is the only interviewer and you get one report per candidate. InstaDataAssess runs the whole funnel, adding live expert-panel rounds on video and a combined AI-plus-expert decision. If you need human panel rounds too, choose InstaDataAssess.
What happens during an interview?
After a short briefing the AI moves through profile, skills, coding, follow-up and probing phases, about 30 minutes by default. It is hands-free voice: candidates can interrupt, and the AI adapts each next question to the skill it knows least about. Interviews can run in English, Spanish, French, German, Hindi, Portuguese, Arabic, Chinese or Japanese.
How does the coding round work?
Python and JavaScript run directly in the candidate's browser, so there is nothing to install. In agentic mode the candidate may use AI tools but must write a justification, which is scored alongside the code. Additional languages can be run server-side on request.
How does it deal with AI-written or memorised answers?
Answers that look AI-generated or read out are always followed by a "defend your answer" probe. If a polished answer gets a weak defence, the AI-likelihood signal rises and feeds the explainable integrity index, alongside in-browser proctoring events such as tab switches, extra faces or paste bursts.
Is photo-ID verification required?
It is optional and set per job. When on, the candidate shows a PAN, Aadhaar, passport or driving licence; the face match runs in the browser and must reach 70% before the interview starts. The document is kept for real interviews and never kept in the public demo.
Does the AI make the hiring decision?
No — it recommends. HR approves the interview blueprint before anyone is invited, reviewers can override any recommendation with an audit trail, and candidates can request human review through an appeal. A model card and four-fifths adverse-impact monitoring are included for your compliance team.
Does it analyse facial expressions?
Facial-emotion inference is not used in scoring or proctoring. There is an optional facial-expression signal for the behavioural profile, but it is off by default, requires the candidate's explicit consent, runs in the browser, stores only an aggregate and is blocked for EU candidates.
How much does it cost?
Free is $0 for 5 interviews a month. Starter is $99 a month for 50 interviews, Growth $299 for 200, and Scale $799 for 1,000. Every plan starts with a 14-day free trial and no credit card. For higher volumes or a dedicated deployment, contact us.
