82%
in-store behaviourDiscovery, hesitation and abandonment leave no trace in today's retail stack.
Measure
everything before the sale.
Turn existing store cameras into a behavioral sensor network. Movement, dwell, intent and emotion quantified at the edge. Zero identity stored.
POS sees one frame of a much longer story.
Online retail measures every click, scroll and abandoned cart. Offline retail measures receipts. Everything between the entrance and the till has been invisible for decades.
1 / 4
trial-room visitsNever convert to purchase, with zero signal on which SKUs drop off behind the curtain.
$603M
India analytics, 2025Growing at 18.7% CAGR to $2.96B by 2034 as brands close the offline measurement gap.
130+
new stores / monthOpening across India in late 2025 — each launching without behavioural telemetry.
Edge inference behind every camera you already own.
One compute unit per store. Existing CCTV feeds. Pose estimation, gaze tracking and affect classification run locally. Raw frames never leave the device.
Ingest camera feeds
RTSP streams from existing CCTV ingested at the edge. No new sensors, no floor hardware.
Extract behavioral signal
3D skeleton reconstruction, trajectory mapping and zone dwell computed per frame. Identities reduced to anonymous vectors.
Model emotion and intent
Affect classification and interaction scoring produce a continuous behavioral state per visitor.
The instrument panel for the physical store.
Every signal a category manager, ops lead or loss prevention team needs — measured continuously, reconciled with POS, surfaced where you already work.
01
Zone heatmaps
Footfall and dwell at shelf, fixture and display level. See which collections pulled attention and which got skipped.
02
Pickup & return
Which SKUs were physically handled, returned to shelf, or carried toward the till. Quantify intent before it becomes a sale or a walk out.
03
Trial-room conversion
Entries, exits, time inside and follow through to purchase. Identify the SKUs that lose customers behind the curtain.
04
Staff coverage
Staff position relative to live shopper demand. Flag understaffed zones during peak intent windows.
05
Loss prevention
Concealment patterns, unscanned exits and collusion detected in real time. Alerts to device, not a tape review next week.
06
Cohort & layout audit
A/B test fixtures, planograms and pricing zones with proper attribution. Merchandising decisions on evidence, not intuition.
30 day pilot. One store.
Installed in under four hours. First behavioral report by day fourteen. No camera replacement. No customer disruption.
TALK TO OUR CEO