Duxbury Networking has introduced the Ironlink AI NVR lineup with CVEDIA Analytics for the South African market, integrated with Milestone XProtect to help partners add edge AI to new or existing video estates without replacing cameras.

Built on enterprise-grade Velasea hardware, Ironlink pairs hardware acceleration at the edge with CVEDIA’s video analytics and Milestone’s VMS workflows. Each unit ships with Milestone preloaded and license-ready, includes a perpetual CVEDIA-RT license (four streams, expandable), and enables remote device management for monitoring and updates.

“The strongest value for partners is practical: add intelligence now, keep the cameras you already have, and scale on your terms,” says Leandro da Cunha, surveillance business unit lead at Duxbury Networking. “Because the analytics run on the NVR with dedicated AI acceleration, customers get real-time detection and cleaner investigations without pushing everything to the cloud or overhauling their network.”

Ironlink systems are powered by Hailo edge AI (with optional NVIDIA GPU configurations), so analytics execute locally, reducing latency and bandwidth while keeping VMS workflows inside the XProtect Smart Client. Within XProtect, operators configure zones (for example, intrusion or loitering), view detections in real time, and conduct attribute-based searches by person or vehicle metadata, directly within the familiar client interface.

The Ironlink portfolio spans 1U 4-Bay units for small edge sites, 2U 8-/12-Bay for medium facilities, and 2U 28-Bay for large or multi-site enterprises. All these are running the same analytics and management stack. Systems include Canopy remote device management, enabling firmware updates and health checks without dispatching field teams.

“For the channel, this solves two common pain points in one box: dependable recording and credible AI,” says Da Cunha. “Partners can modernise a site simply by upgrading the head-end (Milestone plus CVEDIA on Ironlink) and then phase camera upgrades over time. It is a low-disruption path to AI that customers can justify and support.”