Retail stores
Ask why a store dipped, get the recap, and see the forecast before the next peak.
Built into Ariadne, powered by your footfall data
Ask your footfall data anythingwith the Threa AI assistant
Threa AI is the assistant inside Ariadne. It reads the same anonymous people counting data your dashboards already run on, and answers in plain language: what happened, what changed, and what is likely next. It can also run automations, so an answer becomes an action.
ConversationalSummaries and reportsAnomaly alertsForecastsAutomationsMultilingualNo PII at capture
Trusted by Global Industry Leaders
What Threa AI is
What it is
Threa AI is Ariadne's built-in assistant for crowd analytics. Instead of building a report, you ask a question. Threa AI reads your live people counting data, the footfall, dwell, and occupancy signals captured across your sites, and gives you the answer, the summary, the alert, or the forecast you were going to build by hand.
Threa AI runs on the same data as the Ariadne Analytics dashboard. Nothing new is collected and no personal data is involved, because Ariadne counts people without identifying anyone.
What you can ask it to do
Five things Threa AI does with your footfall data
Threa AI is not a chatbot bolted onto a marketing site. It works on your real people counting data, inside the product, with five jobs.
ask
01
Ask in plain language
Type a question the way you would ask an analyst. How did the weekend compare to last weekend across all stores? Threa AI reads the footfall, dwell, and occupancy data and answers, with the chart if you want it.
summary
02
Summaries and reports
Threa AI writes the recap you used to assemble by hand: a daily site summary, a shift handover, a weekly KPI brief. The numbers come straight from your people counting data, not from a guess.
alert
03
Anomaly alerts
Threa AI watches your patterns and tells you when something is off, an unusual spike, a drop in dwell, occupancy creeping toward a limit, so you hear about it instead of finding it later in a dashboard.
forecast
04
Forecasts
Threa AI projects what is likely next from your own history and live signal, so staffing decisions, opening hours, and space planning work to the curve ahead instead of reacting to yesterday.
automation
05
Run automations
Turn an answer into an action. When a pattern repeats or a threshold is crossed, Threa AI can trigger the follow-up you set up, so the insight does not just sit in a dashboard waiting to be noticed.
The data underneath
Threa AI reads counting data, never personal data
Threa AI answers are only as trustworthy as the data beneath them. That data is Ariadne's: footfall counted by Hybrid Fusion sensors that measure presence and movement without cameras and without collecting any personal identifier at the point of capture. Threa AI reasons over those anonymous counts. It does not add a single new piece of tracking.
No cameras. No facial recognition. No PII at capture.
The assistant inherits the privacy posture of the platform.
See the people counting platformAsking vs building
What changes when the report writes itself
Most teams already have the data. The cost is the time between a question and the answer.
The alternative
Build it by hand
Dashboards, exports, spreadsheets
With Ariadne
Ask Threa AI
Plain-language, on your live data
Getting an answer
Open the dashboard, filter, export, chart it.
Getting an answer
Ask the question. The answer comes back with the data behind it.
Recaps and handovers
Someone assembles the summary every shift.
Recaps and handovers
Threa AI drafts the recap from the same numbers, on schedule.
Catching problems
You find the anomaly when you next look.
Catching problems
Threa AI flags the anomaly when it happens.
Looking ahead
A forecast is a separate analysis project.
Looking ahead
Threa AI projects from your history and live signal on request.
The ground truth Threa AI runs on
Threa AI is only as good as the data under it, so the data is the point
Threa AI does not invent insight. It reads the same anonymous footfall record Ariadne has been building across the physical world.
700M+
Anonymous trajectories
Footfall paths Ariadne has captured to date, the record Threa AI reasons over.
32
Countries
Where Ariadne's people counting runs today.
12K+
Devices
Sensors feeding the live signal Threa AI reads.
3.5M+
Square metres
Floor space measured across deployments.
Where teams ask the most
Threa AI earns its keep wherever footfall drives the decision
Shopping centers
Compare tenants and zones in plain language, without building a report per question.
Airports
Surface queue and dwell anomalies across terminals the moment they appear.
FAQ
Frequently asked questions about Threa AI
What is Threa AI?
Threa AI is the assistant built into Ariadne. It answers questions about your people counting data in plain language, writes summaries and reports, flags anomalies, forecasts what is likely next, and can run automations so an answer becomes an action.
Is Threa AI a separate product or part of Ariadne?
Threa AI is a feature of the Ariadne product. It runs inside the same place your analytics already live, on the same data.
What data can Threa AI see?
Threa AI reads your anonymous footfall, dwell, and occupancy data, the counts captured by Ariadne sensors. It has no access to personal data, because none is collected at capture.
Does using Threa AI add any new tracking or collect personal data?
No. Threa AI reasons over the counts the platform already produces. No cameras, no facial recognition, no PII at capture.
What kinds of questions can I ask?
Comparisons over time, site and zone breakdowns, shift and daily recaps, anomaly explanations, and short forecasts. If the answer is in your footfall data, Threa AI can reach it.
Does Threa AI reply in more than one language?
Yes. Threa AI understands and answers in multiple languages, so teams across regions can ask in the language they work in.
Resources
Read what's working
Field reports from real customers, technical deep-dives on how Hybrid Fusion actually behaves, and the occasional opinion piece.
Blog
What Is an Anchor Tenant? Definition, Examples, Economics, 2025 Decline
Anchor tenants are the large retailers that drive footfall to a mall. Definition, types, lease economics, and what is replacing department stores in Europe.
Blog
V-Count alternative: how Ariadne's camera-free stack compares
V-Count alternative compared: sensor type, accuracy, demographics, privacy, deployment, geography, pricing. Built from public docs with sources cited inline.
Blog
Time-of-Flight sensor explained: how a ToF people counter works without a camera
How a Time-of-Flight (ToF) people counter works: depth maps not images, lighting-independent accuracy, and Ariadne's ToFu as a camera-free worked example.
Talk to us
Two questions, twenty minutes, a real walkthrough of your venue's footfall.
What to expect
- 20-minute screen share, walked through on your venue map
- Live walkthrough of Hybrid Fusion sensor outputs
- Where Ariadne fits, and where it doesn't
Got a different question?
Send us a message
Anything that isn't a sales conversation. We'll route it to the right person and get back within one business day.