Skip to main content
Proctorly runs monitoring in the browser for the length of an exam. There’s no extension and nothing to install. When a student grants camera and screen-share access during setup, monitoring begins and continues until they finish.

​
What Proctorly watches for

Depending on what the teacher turns on, Proctorly can watch for:

Face presence

Flags when no face is visible, and when more than one person is visible.

Test focus

Flags leaving the exam tab or switching windows, and leaving full-screen mode.

Environment

Flags loud or suspicious background audio, and can capture a few random snapshots for later review.

Screen sharing

Requires sharing the entire screen; flags when sharing is interrupted, and detects additional displays.
The full list of options — and what each one catches — is in Configuring monitoring.

​
The student session, end to end

A student walks through a short, guided flow. Steps in brackets only appear if the teacher enabled them.
1

Consent

The student sees what’s monitored, enters their name and email, and agrees to be monitored.
2

[ID verification]

If enabled, the student photographs their student ID.
3

Setup

The student grants camera and screen-share access; face detection loads.
4

[Waiting room]

If enabled, the student waits until the teacher admits them.
5

Quiz

Monitoring is active. The student completes the embedded exam content.
6

Done

The student submits, the session ends, and they see a completion message.
If a student reloads mid-exam in the same browser tab, Proctorly resumes them at the step they left off rather than restarting.

​
What the teacher reviews

After sessions finish, the Report screen shows each student’s session with an integrity score (starting at 100, with points deducted per flag) and the flags detected — each with a confidence and a risk level. The teacher confirms or dismisses flags, can mark a session Clear or Escalate, reviews snapshots, and can export to CSV. See Reviewing results.
Proctorly can generate an AI-written narrative report and summary emails, but these are assistive only. The flags themselves are detected by the monitoring engine, not by AI.
Last modified on June 24, 2026