QuantaTemp logo QuantaTemp public system map

Interactive Architecture

How QuantaTemp turns weather into guarded market decisions

A read-only map of the runtime loop: data collection, model scoring, contract selection, live execution, safeguards, and the feedback systems that tune the next cycle.

9 cities T+0 - T+4 horizons 30m forecast cadence 15m market cadence 10m fill polling
QuantaTemp architecture diagram Interactive SVG showing inputs, scheduler, database, model pipeline, execution path, safeguards, and feedback loop. Live inputs App and scheduler Persistent state Model and selection Execution Forecast modelsOpen-Meteo snapshots30 minute cadence Kalshi marketsquotes, books, fills15 minute snapshots NWS observationsstation temperaturesintraday blending Official actualsfinal highs and lows3-9 AM local polling Dashboard + APIsFlask server.pyHTML, JSON, SSE logs Manual controlsrun, reconcile, drainconfig and safe mode Enhanced scheduleradaptive job cadencefeeds, cycles, fills Collector layerforecast, market, obsnormalized writes Safeguardspreflight + drawdownfrozen-config gates Raw data tablesweather + market rowslatest and history Database runtimePostgres productionSQLite local mode Model statesignals, weightscalibration, ML stats Trade statepaper + live ledgersconfig and orders Telemetryworkflow + executionrejections and shadows Model cycleT+0 to T+4 horizons30/60 min adaptive Ensemble enginebias + uncertaintysignal confidence Contract pricingprobability + EVfill and fee adjusted Market universeallowed/downweightedblocked segments Portfolio selectioncorrelation + caps75% Kelly sizing Paper pathsimulated entriesvalidation ledger Live entry gatequality + risk gatesdual-side, drawdown Orders + fillsstaged executionrepricing + MTM Exits + reconcileclose or verifyauto state repair Analytics APIdiagnostics + attribution Learning loopweights + calibration