01 · Utilisation
“How was X used last week?”
Hour-by-hour, day-by-day, with the 4-week average overlay so you can see the deviation, not just the number.
01 · Technology · Talk to your data
Skip the spreadsheets. Just ask the building.
An ops lead shouldn't need a SQL course to find out how the library was used last Friday. Ask in plain English, the assistant queries your anonymous occupancy data, runs the math, and answers in a chart, a number, or a one-line summary.
02 · How a question becomes an answer
Four steps. About six seconds.
No dashboards to open. No filters to drag. No “let me check with the analyst on Monday”.
01 · You ask
How was Library Floor 4 used last Friday?
03 · Try the demo
Ask anything. Or click a suggestion.
The demo runs against simulated data from a 5-zone university library. In production this is wired to your live deployment.
04 · What you can ask
Eight questions worth asking.
Utilisation, comparison, safety, scheduling, energy, reporting, alerting, export: one assistant, one way to ask.
02 · Comparison
“Compare exam week to teaching week.”
Side-by-side. Per zone. With a calm summary of the three differences that actually matter.
03 · Safety
“Did we cross density thresholds?”
Threshold crossings with timestamps, an egress estimate, and a one-line incident note for the safety log.
04 · Scheduling
“When should we open this space?”
Suggested opening hours based on the last N weeks of actual demand. Saturday morning, every time, is the surprise.
05 · Energy
“Where did we waste HVAC?”
Zones running at peak schedule against <20% occupancy. Annualised AUD impact next to each row.
06 · Reporting
“Write me the Q1 sustainability report.”
Draft PDF + raw CSV. Methodology, baseline, savings, formatted for the NABERS submission pack.
07 · Alerting
“Tell me when atrium > 90%.”
The assistant turns the request into an alert rule with the right calm-period, and routes it to your channel of choice.
08 · Export
“Email this to finance every Monday.”
Scheduled exports of any query result: CSV, PDF, or pushed direct into Snowflake / BigQuery.
The privacy answer hasn't changed.
The assistant only sees what the platform sees: anonymous counts.
There is no video to query. There are no faces to look up. The model has access to the same aggregate counts the dashboard does, nothing more. The architecture remains the privacy answer.
- No video access, ever
- No biometric, no PII in the prompt context
- Conversation history scoped per-tenant
- Every response cites the underlying counts
05 · Bring a question
A demo with your floorplan answers more in twenty minutes than a brochure in twenty pages.
Bring the question you've been wanting to ask. We'll wire the assistant to a sample of your data and have it answer in the demo.