Computer vision workspace

See the feed. Name the face. Gate the door.

Gnosis is the control plane for live face pipelines: profiles and groups in one place, detection or recognition on the recognizer, signaling through the main API.

Modes
Detect / ID
Origins
Cam / device
Graph
3 hops
Live feed · webcam
MATCH 98%
Ava Chen · enrolled
1920 × 1080processor · recognition

origin

webcam

processor

recognizer

destination

overlay

Workspace

Four surfaces, one pipeline

Find

Face detection

Lock onto faces in a still or a live camera feed — the same Play / Stop path you already use in the workspace.

Match

Face recognition

Build encodings from profiles, then identify people against a group dataset in video.

Organize

Profiles & groups

People records, reference images, and datasets for recognition, vigilance, and access lists.

Route

Live pipeline

Origin, processor, and destination as an explicit graph — webcam or device into the recognizer and back.

How it runs

From enrollment to a live overlay

  1. 01

    Enroll

    Create profiles, attach reference images, and group the people you care about.

  2. 02

    Wire the feed

    Point a browser camera or a device origin at detection or recognition.

  3. 03

    Act on matches

    Gate access, watch a live overlay, and keep the pipeline explicit end to end.

Spin up a workspace and point a camera at it.

Sign in, enroll a profile, pick a group, and run detection or recognition from the browser.