CertiVue analyses CCTV alarm clips in under 10 seconds and returns a confidence score that separates genuine security events from false triggers. It also adds fire and smoke detection, plus a Situation Room where operators can generate, test, and review realistic scenarios before they happen.
Trusted by security professionals
Situation Room gives teams a controlled way to generate CCTV scenarios from live camera views, run detection against those generated clips, and compare review outcomes before incidents happen in the real world.
Create synthetic clips directly from the camera viewpoint to test intrusions, vehicle movement, and emergency response workflows.
Review how custom fire and smoke models respond alongside people and vehicle detection, with annotated runs grouped under each generated clip.
Operators can compare generated source clips, expanded detection runs, and review outcomes in one place to improve confidence before go-live.
Every false alarm drains resources, erodes trust, and puts genuine threats at risk of being overlooked.
Operators spend hours reviewing footage that turns out to be wind, rain, or shadows. Every false review is time stolen from genuine threats.
Constant false alarms desensitise response teams. When every alert feels like crying wolf, real incidents risk being deprioritised.
Customers churn when they lose confidence in monitoring quality. Police deprioritise repeat false alarm sites, increasing liability.
CertiVue solves this at the source — before the alarm ever reaches an operator.
CertiVue slots seamlessly into your existing Sentinel workflow. No changes to your alarm handling process.
Your Sentinel system detects camera motion and captures a video clip automatically.
CertiVue's AI models process the clip, detecting people, vehicles, fire, and more — including custom-trained models for specialist threats.
Detections are filtered against your configured inclusion and exclusion zones to eliminate noise.
A probability score (0-100) is returned to Sentinel, letting operators prioritise real threats.
Powered by state-of-the-art AI with dedicated people, vehicle, fire, and smoke detection — plus custom-trained models for specialist security and safety threats.
Identify human presence with high confidence, filtering out animals and environmental movement.
Cars, vans, trucks, and motorcycles. With smart stationary vehicle filtering to ignore parked vehicles.
Identify visible flame events early so operators can escalate genuine fire incidents immediately.
Catch early smoke build-up before open flame is visible, improving response time for high-risk indoor and industrial sites.
Water ingress and flooding alerts to protect assets and trigger evacuation procedures early.
Detect doors left open or forced entry points, critical for perimeter security monitoring.
Our architecture supports any object type the model can detect. Custom detection types available on request.
Draw inclusion zones to focus detection on entry points. Add exclusion zones to ignore trees, roads, or areas with constant movement. Toggle detection types per zone.
A purpose-built interface for reviewing events, managing sites, and tracking false alarm reduction across your entire portfolio.
Review events with annotated video showing exactly what was detected and where.
Track your reduction percentage over time with interactive charts and trend data.
Manage your entire portfolio in a structured, drill-down interface with per-camera configuration.
CertiVue connects directly to your Sentinel alarm receiving centre. No changes to your existing workflow.
Your existing alarm receiving centre sends video clips when cameras trigger.
Analyses video, applies zone filtering, and returns a probability score in under 10 seconds.
Operators see prioritised events with confidence scores and annotated video evidence.
Customers, sites, and cameras sync automatically from Sentinel. Zero manual setup.
Every alarm is analysed in real-time with a guaranteed sub-10 second response.
Each event receives a 0-100 confidence score that flows directly into Sentinel's workflow.
Valid alarms include AI-annotated video with bounding boxes for instant operator verification.
Not a generic AI bolted onto a camera. CertiVue is engineered from the ground up for security video analysis.
State-of-the-art object detection combined with custom-trained models for fire, flood, and open door scenarios.
Real-time processing on dedicated NVIDIA GPU infrastructure for consistent sub-second frame analysis.
Guaranteed response time that fits within Sentinel's alarm handling workflow. No delays.
Choose from multiple model sizes to balance speed and accuracy. Switch models at runtime without downtime.
Start small, scale as you grow. Every plan includes a 30-day free trial.
For small security installers getting started with AI monitoring.
For growing businesses with multiple sites and advanced needs.
For large-scale operations requiring dedicated infrastructure.
Book a personalised demo with our team. We'll show you exactly how CertiVue integrates with your setup.