Online proctoring is how institutions supervise remote exams without a physical invigilator in the room. The student's webcam and microphone feed into monitoring software that logs rule violations — extra faces, leaving the frame, switching tabs, background voices — while the exam runs inside your LMS.
Schedule a DemoIn a lecture hall, an invigilator can scan rows, check IDs at the door, and confiscate phones before the clock starts. Online, those controls have to be rebuilt digitally — and they have to work for students sitting in hostels, rural broadband connections, and shared family spaces.
Online proctoring fills that gap. At exam start, monitoring software confirms who is at the device. During the attempt, it watches for behaviours that break exam rules: someone else entering the camera view, the candidate ducking out of frame, browser tabs changing, or voices that suggest off-screen coaching. Each incident is tied to a timestamp so reviewers can judge context instead of guessing from a single snapshot.
The goal is not to replace academic judgement with automation. It is to shrink the monitoring problem from “watch every student for two hours” to “review the handful of sessions where something looked wrong.” That shift is what made 70 postgraduate entrance papers feasible for 12,000 SPPU candidates and 60,000 fellowship sessions manageable for Natview Foundation without staffing a proctor for every screen.
Vendors use overlapping labels — “live,” “automated,” “hybrid” — but the operational difference is who watches the feed and when. Use this table when shortlisting tools; the right model depends on exam stakes, cohort size, and how many reviewers you can assign during peak week.
| Model | Mechanism | Fits when | Scale ceiling | Cost pattern |
|---|---|---|---|---|
| Recorded review | Webcam snapshots or clips are stored for faculty to review after the exam window closes. | You need an audit trail but cannot staff a live control room for every session. | Review queues grow with cohort size — plan reviewer capacity ahead of peak terms. | Moderate; human review time is the main variable. |
| Live human proctoring | A trained proctor watches candidates in real time and can intervene during the exam. | Regulators or accreditors require synchronous human oversight. | Low — each proctor covers a small group at a time. | Highest per candidate; scales linearly with headcount. |
| AI-assisted monitoring | Software watches the webcam and microphone feed, logging timestamped events when behaviour drifts from exam rules. | You run frequent quizzes or entrance exams for thousands of students on modest bandwidth. | High — SPPU ran 2,000+ simultaneous sessions; NFTI processed 60,000 sessions over four months. | Lowest at volume when review focuses on flagged cases only. |
| Hybrid oversight | AI handles continuous monitoring; humans step in only for flagged or high-stakes sessions. | One institution runs both routine module tests and regulated final examinations. | High — automation absorbs volume; humans handle exceptions. | Between AI-only and full live models. |
Feature lists on vendor sites look similar. In deployment, these six capabilities separate tools that integrate cleanly from tools that create a second support queue every exam season.
| Capability | Why it matters | How ProctorLink handles it |
|---|---|---|
| Identity at exam start | Confirms the enrolled candidate is the person at the keyboard before questions appear. | Baseline photo capture at quiz launch; optional ID verification for higher-stakes settings. |
| Face count in frame | A second person on camera is one of the most common integrity violations in remote exams. | AI flags multiple faces with still images tied to exact timestamps for reviewer context. |
| Focus on the exam tab | Leaving the browser tab often means searching for answers or messaging someone. | Tab-switch detection with configurable warnings before an event is logged. |
| Ambient audio | Coaching from off-camera is harder to spot visually but shows up in microphone analysis. | Audio event logging alongside visual flags — no separate proctoring portal required. |
| LMS-native workflow | Standalone proctoring portals add logins, sync delays, and support tickets for students. | Moodle quizaccess plugin — proctoring toggles per quiz inside existing course admin. |
| Data residency | Student images and flagged logs are sensitive; many institutions cannot ship them to vendor clouds. | Proctoring artefacts stay on your Moodle server; ProctorLink does not retain student media. |
Remote oversight is no longer limited to MOOC pilots. These are real deployment patterns from ProctorLink customers:
Across these deployments, ProctorLink has supported more than one million proctored exam sessions. Read the full narratives in our case studies.
Planning remote exams for your next academic term? These pages go deeper on implementation and pricing.
30-minute session mapped to your LMS setup, exam calendar, and cohort size.
Subscription tiers for steady exam cycles and credit packs for seasonal peaks.
SPPU, RT-MSSU, and NFTI — numbers, timelines, and what broke before it worked.
quizaccess_quizproctoring plugin — typical go-live in one to two business days.
Concurrency, faculty review workflows, and integrity at entrance-exam scale.
Proctoring for credential programmes without physical test centres.
ProctorLink plugs into Moodle, keeps proctoring data on your servers, and flags incidents — not students — so faculty stay in control of outcomes.