Know exactly who
played fair
Tab-switch detection, face tracking, and environment monitoring run quietly in the background during every interview. Each candidate gets a trust score so you know exactly who played fair.
99.2%
Fraud caught
<1%
False positive rate
6
Detection layers
0 sec
Detection delay
How proctoring works
Silent, seamless, and thorough. Candidates focus on the interview while Anika monitors integrity.
Interview begins
When a candidate joins the AI interview, proctoring activates automatically. Camera access and browser permissions are requested once — no extra software needed.
Continuous monitoring
Anika monitors tab switches, face presence, eye tracking, audio anomalies, and environment changes throughout the entire interview in real time.
Flags raised quietly
Suspicious activity is logged silently — no interruptions to the candidate. Each flag is timestamped with a screenshot and categorized by severity.
Trust score generated
At the end of the interview, each candidate receives a trust score from 0-100 with a detailed breakdown of any flagged incidents and their severity.
6 layers of detection
Multiple independent signals combine into a single trust score. No single trick can fool the system.
Tab-switch detection
Detects when candidates switch browser tabs or windows during the interview. Every switch is logged with timestamp and duration.
Face tracking
Continuous face presence verification ensures the same person who started the interview is the one answering. Multiple faces trigger an instant flag.
Gaze monitoring
Tracks eye movement patterns to detect if a candidate is reading from another screen or receiving off-camera assistance during responses.
Environment analysis
Monitors the candidate's surroundings for unexpected changes — additional people entering the frame, suspicious objects, or screen reflections.
Identity verification
Compares the interview participant against the candidate's uploaded ID or profile photo to prevent proxy test-taking.
Audio anomaly detection
Listens for whispered prompts, text-to-speech output, or background voices that suggest external assistance during the interview.
Understanding trust scores
Every candidate gets a 0-100 score after their interview. Here's what the ranges mean.
No flags detected. Clean interview.
Minor flags (e.g., brief tab switch). Review recommended.
Multiple flags detected. Manual review required.
Serious integrity concerns. Interview may be invalid.
Common questions
Do candidates know they're being proctored?
Yes. Candidates are informed before the interview starts that proctoring is active. They consent to camera and browser monitoring as part of the interview process. Full transparency.
What if a candidate has a legitimate reason for a tab switch?
Single, brief tab switches (under 3 seconds) are noted but scored with minimal impact. The trust score weighs frequency, duration, and context. A single accidental switch won't tank a score.
Does proctoring work on mobile devices?
Yes. Mobile proctoring uses the front-facing camera and monitors app switching. The experience is optimized for both iOS and Android browsers.
Can I disable proctoring for specific roles?
Yes. Proctoring is configurable per job. You can enable full proctoring, light monitoring (tab switches only), or disable it entirely based on the role and your preferences.