Proctoring

Prevention

Rosalyn’s system reduces opportunities for candidates to cut corners. System lockdown and ID verification remove the temptation for the most common test violations.

ID verification at test start and continuously during test.
Prevent external devices being attached to the computer
Block recording or other applications from running.

Deterrence

Setting expectations of exam integrity dissuades candidates from compromising their own. With Rosalyn, candidates understand how the exam is invigilated and the consequences of misbehavior.

Require test-takers to understand and agree to test rules.
Alert test-takers of problem behavior in real-time.
Present your organization as a vigilant proctor who takes security seriously.

Detection

Human proctors catch only what they can see or hear. Rosalyn’s human-in-the-loop AI analyzes many more datastreams. The multi-attentive AI is alert to novel strategies to violate test rules or capture test content.

Second order 3D models discern gestures, gaze, and movement.
Uncover subtle cues to possible test violations.
Identify attempts to record the test content with cameras and devices

Arbitration

Rosalyn’s human-in-the-loop proctoring solution affords the efficiency and consistency of automation while keeping the decision-making authority in your hands.

Test violations are flagged by the AI based on clear and objective criteria.
Proctors review test session events in real-time.
Sophisticated queuing distributes events efficiently to network of proctors to eliminate bottlenecks and support unlimited scale.

Action

Rosalyn’s system flags issues as they happen. The AI takes action on minor problems by alerting or warning the student. Serious violations are escalated and acted upon by human proctors.

Annotated recordings provide conclusive evidence.
Review only the relevant data from the test session.
Real-time alerts to test-takers keep problems from escalating.

Discovery

No security is perfect. Candidates are always inventing new ways to gain an unfair advantage. Rosalyn’s system is continually learning from test sessions to identify novel threats to the integrity of your assessments.

Speed up discovery with machine learning.
Beyond the catalog of known behaviors, the system flags anomalous patterns which point to potential test violations.
Updated AI models can be put into production immediately.

Forensics & Analytics

Spots evidence of leaked exams with machine learning.

Analytics give you the bird's eye view of your assessment program. Forensics offer the tools to improve your test integrity. Some threats to test integrity like pre-knowledge of test content are discoverable by analyzing test results across demographics, locations, and affinity groups.

Rosalyn uses machine learning on aggregate data to identify patterns not visible in individual test sessions.
Session data correlated to psychometrics can differentiate between test-takers getting outside assistance and those with pre-knowledge of test content.
Know immediately if and when test content becomes compromised.

Superhuman Invigilation

More than just detection, Rosalyn systematically improves the integrity of your exams at every touchpoint.