Forces can forecast demand. They still cannot say what to do tonight.
PivotPoint AI sits between a force's systems and the commander's live deployment: hyper-local demand, the people actually available, a constrained recommendation, and a human decision.
The missing operating layer
Forces can already forecast. PivotPoint sits between those systems and tonight’s deployment decision — then keeps the audit trail.
Forecast demand
Where, when, and what type — with calibrated uncertainty and a labelled time window.
See who can actually go
Roster, skills, licences, location, fatigue, overtime. Synthetic in this demo until a force agreement exists.
Recommend under constraints
A deployment that is feasible tonight, not a heatmap with a suggested pin.
Decide
Accept, modify, or reject, with a reason. The human remains in command.
The workspace commanders already sit in
Below is the live Merseyside demo — public data, synthetic officers, a human still in command. These are product photographs, not concept art.
Forecast sits on the map, not in a slide pack
Monthly Police.uk demand for Merseyside, with forecast geography labelled. Colour is risk percentile within the force — not a probability of any incident.
Who is actually available
Patrol units on the map are synthetic until a force data-processing agreement exists. Roster truth is the point — not a heatmap of where crime might be.
The geofence is the operational area
Drawing is a step in the commander chain, not a hidden map tool. What follows is who is actually available — and what that costs.
The live demo is trained on public Police.uk history for Merseyside (~501,391 events), not live STORM, NICHE, or roster feeds. Resource layers are synthetic by design until a force data-processing agreement exists. PivotPoint will not sell person-level prediction or autonomous dispatch.