How AVENGE Works

A consent-based, privacy-first approach to community video intelligence. From camera to investigation, every step respects your control.

Connect

Any camera works: Ring, Nest, dashcams, RTSP

Process Locally

AI runs on your device, not in the cloud

Control Sharing

You decide what to share, with whom, when

For Camera Owners

The contributor journey

How homeowners and businesses help their communities while maintaining control.

1

Connect Your Cameras

Link your existing security cameras, Ring doorbell, Nest, or dashcam to AVENGE. Our edge device connects via RTSP or cloud APIs.

Works with Ring, Nest, Wyze, Arlo, Eufy, and any ONVIF/RTSP compatible camera. No new hardware required for most setups.

2

Local Processing

Our edge device analyzes your footage locally using AI. It detects vehicles, license plates, and people without sending raw video to the cloud.

Uses YOLOv8 for detection and specialized LPR models. Processing happens on a Raspberry Pi 5 with Hailo-8 accelerator.

3

Receive Investigation Requests

When a crime occurs nearby, you get a notification asking if you might have relevant footage for the specified time window.

Requests include incident type, time window, and approximate location. You see exactly what investigators are looking for.

4

Preview & Approve

Review what you'll share before approving. Toggle individual elements like face data, full video, or just metadata.

You maintain full control. Revoke consent anytime before evidence is exported to the investigation.

5

Earn Rewards

When your footage helps an investigation, you earn contributor points redeemable for premium features.

Points unlock extended cloud storage, advanced analytics, and recognition in community reports.

For Investigators

The investigation workflow

How law enforcement and verified investigators use AVENGE to solve crimes.

Step 1

Create Investigation

Log an incident with location, time window, and details. The system identifies cameras that may have relevant footage.

Works for law enforcement, insurance investigators, and verified community investigators.

Step 2

Request Footage

Send consent requests to camera owners in the relevant area. They decide whether to share.

Owners see your verified credentials and the specific incident you're investigating.

Step 3

AI Analysis

Shared footage is analyzed to detect vehicles, track license plates, and correlate across multiple cameras.

Cross-camera correlation builds complete timelines from multiple contributor sources.

Step 4

Evidence Package

Receive a court-ready evidence package with chain of custody documentation and metadata.

Full audit trail documents consent, access times, and any modifications.

Architecture

Edge-first by design

Processing happens locally. Only metadata travels to the cloud.

Contributor Layer
Your Cameras
Edge Device
Local Storage
Stays on your network
Processing Layer
Object Detection
LPR Engine
Cross-Camera Correlation
Metadata only
Application Layer
Investigation Map
Timeline View
Audit Trail
Consent-gated access
Technical Details

Under the hood

Built on proven open-source technologies with enterprise-grade security.

Edge Hardware

Processor
Raspberry Pi 5 or Jetson Orin Nano
AI Accelerator
Hailo-8 (26 TOPS)
Storage
256GB+ SSD
Connectivity
Ethernet + WiFi 6

AI Models

Object Detection
YOLOv8/v9 (Ultralytics)
License Plate
Custom LPR + Plate Recognizer
Tracking
ByteTrack
Re-identification
FastReID with OSNet

Data Security

Encryption
AES-256 at rest, TLS 1.3 in transit
Authentication
OAuth 2.0 + MFA
Audit Logging
Immutable, 7-year retention
Compliance
CCPA, BIPA, SOC 2 (in progress)

Ready to get started?

Join the waitlist for early access to AVENGE. Be among the first to help build a safer, more private community.