Deploy intelligence wherever it belongs
One analytics stack, three deployment planes — mix them freely across a single estate, with structured events flowing to your command centre.
Three planes, one stack
Edge
Models run on-camera or on rugged edge boxes — sub-second alerts even with zero connectivity.
On-Premise
GPU servers inside your data centre for full control, compliance and air-gapped networks.
Cloud
Video cloud software for elastic, multi-site estates — central dashboards, updates and storage.
Plugs into the systems you already run
Camera-agnostic ingest and an API-first event pipeline.
ONVIF & RTSP Ingest
Discover and stream from existing IP cameras and NVRs — no rip-and-replace.
CAMERASREST API & Webhooks
Push events, metadata and alerts into any dashboard, SIEM or PSIM.
APIVMS & C4I Bridge
Surface Videolytical analytics inside your current VMS or command-centre console.
INTEROPSMS / Email Gateways
Alert routing to phones and inboxes with escalation rules and schedules.
NOTIFYAD / LDAP & RBAC
Enterprise sign-on, role-based access and per-camera permission scopes.
IAMAir-Gap & Offline Mode
Full operation on isolated networks with signed, offline updates.
DEFENSE-READYSecurity & compliance, by default
For on-premise and edge deployments, footage never leaves your premises — analytics run where the cameras are, and only structured events cross the network. Transport is encrypted end to end, every alert and operator action is written to an audit trail, and access is scoped per camera and per function.
- TLS-encrypted streams & APIs
- Full audit trails of alerts & actions
- Role-based access per camera & function
- Signed model & software updates
- Data sovereignty — your video never leaves your network (on-prem)
Deployment checklist
- Site survey & camera audit
- Network & GPU sizing
- Pilot on 4-8 cameras
- Scale-out with calibration
- Handover + training
Questions, answered
Can Videolytical work with our existing CCTV cameras?
Yes. The stack is camera-agnostic and ONVIF-friendly — analytics are added via edge boxes or central GPU servers that ingest your current streams, so there is no rip-and-replace of the camera estate.
Do we need cloud connectivity for the AI to work?
No. Models run fully offline on edge devices or on-premise GPU servers. Cloud is an optional plane for elastic, multi-site management — not a dependency for detection or alerting.
Which analytics modules are available out of the box?
30+ modules including intrusion, tripwire, loitering, crowd & people counting, abandoned/missing object, face recognition, ANPR, helmet & traffic violations, PPE, fire & smoke, heatmaps and video synopsis — plus custom model training for site-specific objects.
How do alerts reach our control room?
In real time into VMS/C4I dashboards, plus SMS, email, third-party VMS or webhooks — with escalation workflows and full audit trails of every alert and action.
Who deploys and supports the system?
Delivered together with OEMs, security solution providers and system integrators — including testing, calibration and ongoing optimization after handover.
Architecture review, on us.
Share your camera count, sites and network — we will come back with a sizing and deployment plan.