Readiness Simulation (RS) Engine
Government-owned mobilization planning simulation for the Selective Service System.
A daily-timestep simulation over aggregate cohorts moving through the six Military Entrance Facility stages: Registration, Selection, Notification, Delivery, Classification and ASP.
This is not a per-registrant microsimulation. PWS 3.0 puts entity-level modeling out of scope and restricts ingest to anonymized aggregate data only. That constraint is what makes the problem tractable: a queueing and flow model with capacity constraints, routing fractions and stochastic rates.
The model, per stage, per day, per iteration
inflow = upstream departures + exogenous arrivals
queue = carryover + inflow
capacity = f(staffing, offices, degraded-operations toggle)
throughput = min(queue, capacity)
losses = throughput x {no-show, medical disqualification, ASP diversion}
departures = throughput - losses
carryover = queue - throughputDeterminism
Same seed, same scenario version, same clock reading gives a byte-identical event stream: order, payloads and CloudEvent ids alike. Re-running a seed reproduces the run exactly, which is how PWS 2.2 transparency, repeatability and after-action reconstruction are met architecturally rather than procedurally.
This deployment
A walking skeleton. It stands up the delivery path end to end, including the server-sent events channel that carries live run progress once the engine core lands. All user interface for this product is asynchronous over SSE. There is no polling.
GET /healthz— livenessGET /api/v1/version— the build being servedGET /api/v1/events/stream— SSE, heartbeat until there are runs to report