WorkSuite — powered by MJC Single stop for day-to-day business

AI Assessment — the right person, fairly scored

Video, voice, written and coding interviews in one AI-run session. Identity is verified before the first question, the session is recorded and locked to the window, gaze and lip movement are monitored for proxies — and every answer is scored against a rubric your panel can see.

4
Formats in one session
ID
Verified before start
30–60
Minute papers
Easy→Hard
Difficulty you set
worksuite / assess / live-sessionRecording
Integrity signalsLive
IdentityVerified
Window focusLocked
Gaze on screen97%
Lip-syncMatch
Other voicesNone
Second face
Q7 · System design · HardAnswer scoring
Correctness84
Depth72
Communication88
Illustrative session
Identity verifiedConsent-based · before start
Flag for reviewSecond face at 18:42
Video interviewsVoice interviewsWritten answersCoding testsIdentity before startAadhaar offline verificationPAN validationSelfie livenessScreen recordingWindow lockGaze trackingLip-sync analysisSecond-face detectionAI-answer detectionEasy → Hard papers30–60 minute sessionsRubric scoringShareable recordings
Complete AI evaluation

Video, voice, written and coding — in one session

Mix formats in a single paper so one sitting tells you how a candidate thinks, speaks, writes and builds. The AI interviewer asks, listens, follows up and scores; your panel reviews the evidence.

worksuite / assess / video
Tell me about a production incident you led. What did you change afterwards?
We had a payment queue backing up during a sale…
You mentioned a retry storm. How would you stop it happening again at ten times the load?
AI follow-ups adapt to each answer

An interviewer that listens and follows up

  • Structured questions for the role, level and skills you choose
  • Adaptive follow-ups based on what the candidate actually said
  • Timed answers with optional one re-record per question
  • Timestamped transcript — jump straight to any answer in the recording
worksuite / assess / voice
VI
Voice round · Customer support, Hindi + EnglishPhone or browser · 15 min
REC
Role-play · angry customer, delayed refund
Acknowledged the delay and apologised without over-promising. Confirmed the order number, explained the refund timeline clearly and offered a callback. Switched to Hindi when the customer did. Missed checking whether the customer wanted an email confirmation.

Spoken skills, tested by speaking

  • Voice-only rounds for support, sales and operations roles — by browser or phone
  • Role-plays with an AI customer, prospect or colleague
  • Multilingual — English, Hindi, Arabic and other languages you enable
  • Clarity, structure and accuracy scored against a rubric — never accent
worksuite / assess / written
Q4 · Case study · Medium12:00 left

A distributor's receivables have grown 40% while sales grew 10%. List the three things you would check first, and why.

1. Ageing by customer — whether a few large accounts are driving the growth. 2. Credit terms granted this year versus last. 3. Disputed invoices and unapplied receipts sitting in suspense.
Words
46
Paste events
0
AI-text check
Original

Reasoning in their own words

  • Short answers, case studies, MCQs and essays in one paper
  • Paste blocked and logged, with keystroke patterns to spot copied text
  • AI-generated text detection as an integrity signal for review
  • Model answers and rubrics you can edit before the paper goes live
worksuite / assess / coding
# Return the k most frequent order IDs
def top_k(orders, k):
    counts = {}
    for o in orders:
        counts[o] = counts.get(o, 0) + 1
    return sorted(counts, key=counts.get,
                  reverse=True)[:k]
  • Basic cases · 6 / 6Pass
  • Edge cases · empty list, tiesPass
  • Performance · 10⁶ orders · O(n log n)

Real code, really run

  • Live editor in 30+ languages with hidden test cases
  • Correctness, complexity, code quality and edge cases scored automatically
  • Replay of how the code was written — not just the final answer
  • Plagiarism and AI-assisted code signals for the reviewer
Identity before the first question

The interview cannot start until the candidate proves who they are

Every session opens with consent and an identity check. The face verified at the start is the face that has to stay on camera until the end.

Integrity during the interview

Proxy and forged interviews, caught on the record

The screen is recorded, the window is locked, and eye and lip movement are monitored for the signs of a proxy, a helper or a script. Every signal is time-stamped against the recording — evidence for a person to review, never an automatic rejection.

01
Screen and session recording
Camera, audio and screen captured for the whole session, with a timestamped transcript.
02
Window lock
Leaving the interview window pauses the session; after the exits you allow, it ends. Copy, paste and new tabs are blocked.
03
Gaze tracking
Repeated looking away from the screen — towards notes or another person — is flagged with the moment it happened.
04
Lip-sync analysis
Speech that does not match the candidate's lip movement — a hidden speaker or dubbed audio — raises a flag.
05
Faces, voices and devices
A second face, another voice, a phone in frame, virtual cameras, remote-desktop and screen-overlay tools.
06
Continuous face match
The face from the identity check is re-matched throughout, so the person who starts is the person who finishes.
worksuite / assess / session-4471 / integrity
Gaze on screen
96%
Window exits
1/ 3 allowed
Flags to review
2
Gaze on screen through the session% per minute
  1. Window exit · 3 sSession paused, warning shown
  2. Gaze away · 38 sLooking left, off screen — reviewer to check
  3. Second face in frameFlag — recording bookmarked
Illustrative session
Our ruleNo emotion recognition

WorkSuite never infers emotions, personality or honesty from a face or a voice. Conduct is reported as observable events.

Our ruleFlags go to a person

Integrity signals are evidence with a time stamp and a clip. A reviewer decides; the system never auto-rejects.

Our ruleConfigured per country

Biometric checks, recording and retention follow local law, with notices and consent shown to the candidate.

Interactive · build a paper

Set the paper — easy to hard, 30 minutes to an hour

Choose formats, difficulty and duration. WorkSuite balances the time and the number of questions — try it below.

Paper settings

Changes update the paper on the right.

Difficulty
Sections
worksuite / assess / paper-builder

Your paper will appear here.

Results

A scorecard your panel can check line by line

Minutes after the session, the hiring team gets the overall score, a breakdown by competency against the role benchmark, the integrity summary and a recommendation — each score linked to the answer that earned it.

  • Rubric scores per answer — correctness, depth, structure and clarity
  • Communication and conduct from what was said and what happened, not from faces
  • Benchmark against the role and against other candidates for the same job
  • Recording, transcript and code replay one click from every score
  • Candidate feedback — strengths and areas to improve, if you choose to share it
worksuite / assess / rohan-t / scorecard
Strong yes
Top 15% for this role
  • Identity verifiedAadhaar QR
  • IntegrityClear
  • Coding91 / 100
Illustrative scorecard
Share with clients & partners

Send the evidence, keep control of it

Agencies, RPO teams and hiring partners can share an interview the way their clients want to see it — the full recording or a one-page scorecard — without losing track of who watched what.

Time-limited links

Links expire when you choose and can be revoked at any moment.

Viewer log

Every view, by whom and for how long, recorded on the candidate.

Candidate consent

Sharing with a named client is checked against the candidate's consent first.

Masked identity option

Share scores and answers with contact details hidden, matching your masked CVs.

Everything in AI Assessment

The complete feature set, filtered your way

Video interviews

AI-led, with adaptive follow-ups and timestamped transcripts.

Voice interviews

Browser or phone, with role-plays and multiple languages.

Written questions

Short answers, case studies, MCQs and essays with model answers.

Coding tests

30+ languages, hidden tests, complexity scoring and code replay.

Easy to hard, or adaptive

Fixed difficulty mixes or questions that adapt to each answer.

Skills-mapped question banks

Your own questions, ours, or AI-generated for the role and reviewed before use.

Aadhaar offline verification

QR code, DigiLocker or Aadhaar app, with consent and no stored Aadhaar number.

PAN validation

Through authorised verification partners, matched to the candidate's name.

Selfie liveness and face match

Live selfie matched to the ID photo, then re-matched throughout.

Other government IDs

Passport, driving licence or national ID for candidates outside India or by choice.

Screen recording

Camera, audio and screen for the full session.

Window lock

Leaving pauses the session; repeated exits end it. Copy, paste and new tabs blocked.

Gaze tracking

Looking away from the screen flagged with the time and a clip.

Lip-sync analysis

Speech that does not match the lips — a hidden speaker or dubbed audio — flagged.

Second face and voice

Other people in frame or in the room raise a flag for review.

Virtual camera and overlay detection

Virtual cameras, remote desktops and on-screen AI helper tools detected.

AI-generated answer signals

Written and coding answers checked for AI-assisted content, as evidence for review.

Rubric scoring

Every answer scored against criteria your panel can read and adjust.

Role benchmarks

Compare against the role benchmark and other candidates for the same job.

Client and partner sharing

Time-limited links with a viewer log, the full recording or a scorecard.

Built into the ATS

Sent from the pipeline; results land on the candidate record.

Connected

Assessment evidence that follows the candidate

Scores, recordings and identity checks live on the candidate record — not in a separate testing tool.

AI Assessment → ATSOne click from the pipeline

Recruiters send a paper from the candidate card; results come back to the same card.

AI Assessment → ClientsEvidence for every submission

Agencies attach scores and recordings to masked submissions, with access logged.

AI Assessment → HRMSIdentity checked once

The verified identity and photo carry through to onboarding and attendance.

FAQ

Questions buyers ask us

How is the candidate’s identity verified before the interview?
Before the first question, the candidate gives consent and verifies their identity: Aadhaar offline verification (QR code, DigiLocker or the Aadhaar app), PAN validation through authorised verification partners, or another government ID where they prefer. A live selfie with liveness detection is then matched to the ID photo, and the same face is re-checked throughout the session. Full Aadhaar numbers are never stored.
What happens if the candidate switches tabs or windows?
The interview runs locked to its window. Leaving it pauses the session, shows a warning and records the event with a time stamp. You set how many exits are allowed before the session ends. Every event is visible to the reviewer with the recording at that moment.
How does gaze and lip-movement monitoring work?
The camera feed is analysed for where the candidate is looking and whether lip movement matches the audio. Repeated looking away, a second face, another voice or speech that does not match the lips raise an integrity flag at that time stamp. Flags are evidence for a person to review — the system never rejects a candidate on its own.
What does “behaviour” scoring mean? Do you analyse emotions?
No. WorkSuite does not infer emotions, personality or honesty from faces or voices. Answers are scored against a structured rubric — correctness, depth, structure and clarity of communication — and conduct during the session is reported as integrity events such as tab exits or a second face. This keeps the assessment explainable and usable under AI-in-hiring rules in the EU, US states and elsewhere.
Can we share interviews with clients and partners?
Yes. Results, recordings and transcripts can be shared through time-limited links with access logged per viewer, and only with the candidate’s consent. You can share the full recording or a summary scorecard.
Can we set the difficulty and length?
You choose the formats (video, voice, written, coding), a difficulty from easy to hard — or adaptive, where questions get harder as the candidate does well — and a duration from 30 to 60 minutes. Question banks can be your own, ours, or generated for the role and reviewed before use.
Book a demo

See AI Assessment on your own data

A 45-minute working session with a WorkSuite specialist — not a slide deck. Tell us how you run today and we will show the same workflow in the product.

  1. Take the interview yourselfSit a short AI interview from the candidate side — ID check to final question.
  2. Build a paper for your roleChoose formats, difficulty and duration for one of your open roles.
  3. Read the scorecardRubric scores, the recording, the transcript and any integrity flags.
  4. Share it with a clientSend a time-limited, access-logged link — the way your clients will see it.
ISO 27001 certified company Reply within one working day No obligation

Book your AI Assessment demo

Pick a slot that suits you. We will confirm it by email.

Preferred time

We use your details only to arrange this demo. See our Privacy Policy.

One platform

The rest of WorkSuite shares the same records

Every module reads and writes one database — so nothing is exported, re-keyed or reconciled between them.

Know who took the interview — and how they did

Sit an AI interview yourself in a 45-minute session, then build a paper for one of your open roles.

Book a demo