View API Specs
Architecture & Infrastructure

Edge-Processed Spatial Intelligence

Aera deploys highly optimized computer vision models directly to on-site NVR hardware. We process high-volume video streams locally to eliminate cloud bandwidth costs and ensure compliance.

0 GB

Raw video uploaded

100%

Local edge processing

Data Flow

The Edge Processing Sequence

01
02
03
04

Local Capture

Inference Engine

Anonymization

Metadata Sync

Existing RTSP camera streams route directly to local NVRs or edge AI appliances over secure local networks.

Highly optimized YOLO and custom spatial models analyze frames locally at 30 frames per second.

Faces and license plates are scrubbed at the hardware level. Only vector coordinates and counts are retained.

Lightweight JSON payloads transmit anonymized coordinates to the cloud dashboard via secure TLS tunnels.

Hardware Specs

Enterprise Hardware Specifications

Our edge architecture runs efficiently on standard enterprise infrastructure, requiring minimal local hardware footprint while delivering high-throughput real-time spatial inference.

NVR Compatibility

Inference Speed

Zero-Leak Protocol

Compatible with standard Intel and NVIDIA NVRs, enterprise Linux OS distributions, and all modern ONVIF-compliant IP cameras.

Delivers sub-50ms inference latency per frame under full multi-camera load, optimizing local processing efficiency without hardware strain.

Features AES-256 local storage encryption paired with secure TLS 1.3 transport protocols for all outbound metadata transmission.

Request Technical Docs

Ready to deploy?

Access our comprehensive technical documentation or schedule an enterprise architecture review with our global engineering team.