AUGMENT VISION

Scaling the future of vision AI

Operate vision AI from one control layer

Deploy, monitor, and scale camera intelligence across sites, regions, and deployment models from a single platform. Augment Vision gives teams one place to connect video, orchestration, and operational oversight without fragmenting their stack.

Learn moreView demo
500K+cameras monitored
30+countries deployed
120+supported camera brands
  • Aramco
  • Thales
  • Quarterhill
  • Medtronic
  • INVISTA
  • ExxonMobil
Times Square1 cameraDublin1 cameraNew Orleans1 cameraFalcon1 cameraNiagara1 cameraNashville1 camera
Sites by location6 sites · 6 cameras
Drag a preview or use its arrow keys to move it independently. Hold Shift for larger steps.
Times Square preview placeholder
Paused
Times SquareUSA1
Dublin preview placeholder
Paused
DublinIreland1
New Orleans preview placeholder
Paused
New OrleansUSA1
Falcon preview placeholder
Paused
FalconAfrica1
Niagara preview placeholder
Paused
NiagaraCanada1
Nashville preview placeholder
Paused
NashvilleUSA1

Hub

Govern every camera, across every location

Organize cameras by tenant, site, and group. Give each user access to exactly what they need, while managing every location from one central Hub.

Filter sites
TenantCity OperationsSites6Cameras6OrganizationLocation based
Deploy a new sitePlay with demo

Build it your way

Record with Agent, store with Vault, and analyze with Hub. Choose where your video lives and how you use it, on-premise or in the cloud.

Record with Agent, store with Vault, analyze with HubFrom left to right, Agent records video from your cameras, Vault receives the recordings and stores them in storage you control, and Hub analyzes recordings with AI workflows, searchable insights, and alerts. Arrows follow the recordings from Agent to Vault to Hub, wherever you deploy the products. Bullet, dome, PTZ, and turret camera styles each connect to their own Agent. Four same-size camera-and-Agent platforms extend to the left, right, and rear of the foreground fleet, fading into the distance to illustrate larger deployments. Vault uses a safe, and Hub sits on a tall pillar with a glass control cabin, a spinning rooftop radar with a sweeping light, and a beacon. Click an Agent device to send a visual pulse along its connected recording lines.Bullet camera with its own AgentAGENTDome camera with its own AgentAGENTPTZ camera with its own AgentAGENTTurret camera with its own AgentAGENTVAULTSTOREDRECORDINGSHUB

Record

Agent

Connect your cameras and record video close to the source.

Store

Vault

Receive recordings and keep them in storage you control.

Analyze

Hub

Turn recordings into insights with AI workflows, search, and alerts.

Extensible and modular at the core

A solid video backbone. A launchpad for your next big idea.

Click a drone or press Enter or Space to take control. Use the arrow keys to fly across the city and up above the clouds. Press Escape, click elsewhere, or move focus away to release control.Click or press Enter to fly. Use arrow keys, then Escape to release.Click or press Enter to fly. Use arrow keys, then Escape to release.

  • Your models

    Your detection. Your inference.

  • MQTT

    Events that reach your devices.

  • MCP

    Video tools for your AI assistants.

  • Custom integrations

    Your APIs and microservices.

  • Workflows

    Your stages. One vision pipeline.

Find every recording with a license plate.

Retrieve recordings tagged "license plate" with a single API request.

Find matching recordings

cURL
curl --fail-with-body --silent --show-error \
  --request POST "${HUB_API_URL}/media" \
  --header "Authorization: Bearer ${HUB_TOKEN}" \
  --header "Accept: application/json; version=2026-01-01" \
  --header "Content-Type: application/json" \
  --data '{
    "filter": { "markers": ["license plate"] },
    "pagination": { "limit": 100 }
  }'

Response / 200 OK

JSON
{
  "data": {
    "media": [
      {
        "id": "507f1f77bcf86cd799439013",
        "deviceKey": "cam-loading-bay",
        "markerNames": [
          "license plate"
        ]
      },
      {
        "id": "507f1f77bcf86cd799439015",
        "deviceKey": "cam-north-gate",
        "markerNames": [
          "license plate"
        ]
      }
    ]
  },
  "metadata": {
    "pagination": {
      "limit": 100,
      "hasMore": false
    }
  }
}

Bring your Vision model. We'll handle the cameras

Wrap any computer-vision task as a stage — ANPR, object detection, quality inspection — and compose it straight into a workflow, right next to the built-in nodes.

Start

Choose how this workflow starts.

Object tracking

Detect and track objects in a recording.

Custom stage

Your own microservice, in any language.

Scene description

Describe a recording with a vision model.

Forwarder

Send results to your own integration.

Layered Enterprise Security

Five layers of security, all protecting your video.

Observability

Hover a layer to explore. Tap or use the keyboard to keep it open.Tap a layer to explore its security controls.

Observability

Every action is audited in detail, so administrators can review activity and report on who did what. OpenTelemetry traces operations across our code and services, with Grafana dashboards bringing together Tempo traces and Prometheus metrics.