1
query language
// data layer · physical spaces
Guest Networks.
Query your building in plain English.
Large venues sit on sensor data from dozens of vendors that can't talk to each other. Guest Networks normalizes WiFi, BLE, and camera signals into one event stream you can actually ask questions of — footfall, dwell time, return rate — in plain English. Private by design, on-prem, no PII stored.
12+
sensor types
99.7%
uptime SLA
on-prem
deployment
what it does
Your building already generates the data. Now you can ask it questions.
Every WiFi AP, beacon, and camera becomes one queryable event stream — no five-dashboard CSV stitching, no SQL.
[01]
Vendor-Agnostic Ingest
APs, cameras, and beacons normalized into one event stream. No per-vendor integration.
[02]
Query in English
AI agents translate "footfall in Zone B at peak" into time-bucketed queries. No SQL.
[03]
Real Metrics
Footfall, dwell time, return rate, zone heat. The numbers that drive decisions, not page views.
[04]
Privacy-First
On-prem by default. Aggregates only, no PII stored. GDPR-ready by design.
how it works
Sense. Ask. Act.
01 · raw signals
Every sensor, one stream
- WiFi APs, BLE beacons, cameras
- Vendor-agnostic ingest
- Dedup + aggregate
→
02 · ask in english
No SQL required
- Plain-English question
- Intent to time-bucketed query
- Answer in seconds
→
03 · act
Decisions, not dashboards
- Bar chart, table, anomaly flags
- Staff + layout decisions
- On-prem, no PII
▸ why it matters — your floor already generates the data. Now you can ask it questions in plain English.
in production
Guest Networks started from one narrow but expensive problem: large venues — malls, stadiums, airports — were sitting on sensor data from dozens of hardware vendors that couldn't talk to each other. An operations manager who wanted footfall in Zone B during peak hours had to export from five dashboards and stitch CSVs by hand. The data was there. The answer was not.
A regional shopping center in the UK running 12 sensor types — three WiFi vendors, two camera systems, BLE beacons, a legacy door counter — connected all of them in a single on-prem deploy. Within a week the ops team ran their first natural-language footfall query and restructured security shifts around real peak hours. The first month saved 18 staff-hours of manual reporting.
Ask your building a question.
On-prem · vendor-agnostic · private by design.