φ(ai) PHI AI
DocumentationDocs
Repository

Scheduling optimization

The clinic's real demand curve — every encounter in the Data Store, bucketed by hour — with overbooking guidance derived from the live no-show model. Counts and arithmetic only: no patient is named on this plane, and no schedule is changed by a model.

AI core: Scheduling optimizer v1.2 · optimization · managed in the Control panel
135,720
encounters behind the demand curve
37
mean no-show risk on the current outreach worklist
2026-08-29 22:54:46
computed · cached one hour

Demand by hour, with overbooking guidance

SlotHistorical volume LoadGuidance
07:00 1,205
book to capacity only
08:00 14,770
overbook +1 per provider-day
09:00 14,693
overbook +1 per provider-day
10:00 14,552
overbook +1 per provider-day
11:00 14,743
overbook +1 per provider-day
12:00 14,786
overbook +1 per provider-day
13:00 14,824
overbook +1 per provider-day
14:00 14,764
overbook +1 per provider-day
15:00 14,461
overbook +1 per provider-day
16:00 14,729
overbook +1 per provider-day
17:00 1,142
book to capacity only
18:00 1,051
book to capacity only

Guidance = no-show probability × slot load × panel factor, rounded — an arithmetic suggestion for a human scheduler, not an automated booking change. The platform holds no write path into any scheduling system from this screen.

SPEC 5.8, live over the synthetic corpus. Overbooking is a policy decision; this screen prices it, people make it.