Skip to content
LocalKVKK-compliantBacked by TSARA Group
How it works

Sees, understands, alerts — all inside your store

The flow from camera to panel; where the heavy lifting happens and why raw footage doesn't go to the cloud.

Three steps

From setup to alert

01

Connect

Connects to existing IP/RTSP cameras; zones (aisle/checkout) are drawn.

02

AI watches

AI detects people and objects and measures behavior anonymously.

03

Instant alert

Events are human-approved, notified and reported.

BEKÇİ AI · Canlı
BEKÇİ AI kamera üzerinde bölge (zone) tanımlama
Architecture

Edge model: your footage doesn't go to the cloud

A small computer (mini-PC) is placed in each store and processes its cameras locally. Only anonymous event data (not footage) goes to the cloud — strong proof of KVKK compliance.

Camera

Existing IP/RTSP cameras — no new hardware needed.

Engine (Edge AI)

AI on a mini-PC: detection, tracking and analysis.

Central panel

All branches report to one secure panel.

  • Minimal bandwidth: only small text data is sent, no video.
  • Reduces the KVKK compliance burden: raw footage stays in your store.
  • Scalable: each store carries its own processing power.
  • Chip-ready: open to acceleration with edge AI chips (Coral, Hailo, Jetson).

Let's adapt the architecture to your store

How many cameras, how many branches? We'll plan the right edge setup for you.