The Stripe for audience analytics

Understand your audience.
Drive real results.

Privacy-aware computer vision turns any camera into real-time audience data — count, demographics, dwell & engagement. We own the analytics engine; you keep your dashboard.

Get started free Go to dashboard View the API
0%
Uptime SLA
0ms
Median API latency
0+
Cameras per tenant
Zero
PII stored at rest
audience.analytics / live Live
In frame
0
▲ 12% vs avg
Viewers
0
▲ 8% vs avg
Avg dwell
0
▲ 5% vs avg
Hourly footfallToday
68% positive mood
Anonymous by design

Built for the operators and networks that already own their screens

Retail Chains Gyms Shopping Malls Airports Restaurants DOOH Networks Digital Signage
Capabilities

Everything you need to measure audiences

One detection engine, every metric your dashboard needs — exposed through a clean, documented REST API and real-time webhooks.

People counting

Accurate footfall, unique visitors and repeat visits with session-only re-identification — never a face stored at rest.

Demographics

Age brackets, gender split, mask & eyewear — each detection scored with a confidence so you can trust the numbers.

Viewing & attention

Head-pose "opportunity to see" tells you who actually looked at the screen — not just who walked past it.

Dwell time

Per-visit dwell and viewing duration, aggregated by minute, hour, day, week and month for instant queries.

Mood & emotion

A trained classifier gives every face a valence score and emotion class — measure how your space actually feels.

Webhooks & alerts

Subscribe to viewer, crowd and daily-summary events. Signed, real-time deliveries straight into your systems.

Pipeline

From camera to API in five steps

The backend is independent of the vision implementation — swap YOLO, DeepStream, OpenVINO or any future edge pipeline without touching your integration.

01

Detect

Any camera feeds the AI detection engine.

02

Analyze

Age, gender, gaze, dwell & mood in real time.

03

Store

Raw events kept immutable; metrics rolled up.

04

Expose

Everything available through the REST API.

05

Visualize

Your dashboard, BI tool or SDK — your choice.

The engine

The data layer that powers dashboards

We don't build another dashboard. We become the analytics engine behind the one you already have.

0%

Query-ready aggregates

Pre-computed hourly, daily & monthly rollups keep dashboard queries instant, even across thousands of cameras.

Anonymous by design

Session-only embeddings, TTL forgetting, zero PII at rest.

Real-time

Live occupancy & viewer streams with webhook push.

Multi-tenant & scalable

Every customer's data fully isolated — from a single screen to a global network.

AI insights

Plain-language daily summaries generated automatically for every location.

Example summary

“1,253 visitors · avg dwell 42s · peak 6 PM. Women viewed 24% longer than men; engagement up 12% week-over-week.”

Use cases

From the shopper journey to the point of sale

Wherever people gather in front of a screen or a storefront, we turn attention into measurable data.

Retail chains

Conversion, footfall by hour and demographic mix per store — benchmark locations against each other.

DOOH & signage

Prove opportunity-to-see and audience demographics per screen for advertisers and network reporting.

Gyms & venues

Live occupancy, peak-hour trends and repeat-visitor rates without a single manual head-count.

Airports & malls

Zone-level crowd analytics and dwell heatmaps across concourses and food courts at scale.

Restaurants

Queue and dwell insight plus mood signals to tune service and layout in real time.

Signage companies

Offer analytics as a value-add to your clients — white-label the data, keep your brand.

Developer-first

API-first. Built for edge devices.

No feature exists only in the UI. Everything the dashboard shows, your code can pull — with keys you manage yourself.

  • REST + Webhooks — read analytics, subscribe to signed real-time events.
  • Python & JS SDKs — zero-dependency clients wrapping the public API.
  • Self-service keys — create, rotate and revoke License IDs with roles.
  • CSV export & OpenAPI — live schema at /docs, exports on demand.
Read the docs Swagger UI 
# Live audience for a location
curl https://api.audience.dev/analytics/live \
  -H "Authorization: Bearer $AA_API_KEY"

# → 200 OK
{
  "in_frame": 47,
  "viewers": 29,
  "avg_dwell_s": 42,
  "occupancy": "busy"
}
from aa_client import AudienceAnalytics

aa = AudienceAnalytics(api_key="aa_live_…")

live = aa.live(location_id="loc_42")
print(live.viewers, live.avg_dwell_s)

# subscribe to real-time viewer events
aa.webhooks.create(
  url="https://you.dev/hook",
  events=["viewer_started", "crowd_alert"],
)
import { AudienceAnalytics } from "aa-client";

const aa = new AudienceAnalytics({ apiKey: "aa_live_…" });

const demo = await aa.demographics({
  location: "loc_42",
  viewersOnly: true,
});

console.log(demo.genderSplit, demo.avgViewerAge);
Pricing

Simple plans. Real quotas.

Start free, scale by cameras and API calls. No dashboard lock-in, no per-seat surprises.

Starter

For a single location testing the waters.

$0/ forever
  • 1 location, 2 cameras
  • Core analytics API
  • 10k API calls / month
  • Community support
Get started Open console
Most popular

Growth

For multi-site chains and networks that scale.

$0.9/ camera / day
  • Unlimited locations
  • Demographics, mood & viewing
  • Webhooks & CSV export
  • 1M API calls / month
  • Priority support
Start free trial Upgrade

Enterprise

For 100+ screens with custom SLAs.

Custom
  • Volume pricing
  • On-prem / edge deployment
  • SSO & audit logs
  • Dedicated success manager
Talk to sales
FAQ

Questions, answered

No. Re-identification uses session-only embeddings that are forgotten after a configurable TTL. We store anonymous analytics events — never images or biometric data at rest.
Not at all. The dashboard is only an admin console. Everything is API-first, so you consume the same endpoints from PowerBI, Tableau, Grafana or your own custom dashboard.
The backend is independent of the vision layer. Any camera that can run the detection service works — YOLO, DeepStream, OpenVINO, NVIDIA or a future edge pipeline all feed the same event API.
A person keeps one identity for a session, so unique counts stay honest. After a long absence the visit number increments — a brief return is the same visit, a later return counts as footfall.
Point a camera at the detection service, drop in your License ID, and events start flowing. Read endpoints and webhooks are available immediately — most integrations ship in a day.

Ready to measure your floor?

Turn every camera into an audience-analytics endpoint. Start free — no dashboard to rip out, no lock-in.

Get started free Go to dashboard Talk to sales
Contact

Let's plan your rollout

Tell us about your screens and we'll map the fastest path from camera to your dashboard.

Talk to saleshello@pixlerlab.com
Implementation helpAPI guide & examples
Security & privacyAnonymous analytics · data protection by design
Thanks — your email draft is ready. Send it when you're set.