Skip to content

Your first forecast run

After data is uploaded and readiness checks pass, trigger a forecast from the workspace or via the API.

Trigger a run (UI)

  1. Open Workspaces and select your workspace.
  2. Click Run forecast (or New run).
  3. Optional Growth+: apply scenario adjustments or config overrides before starting.
  4. You are redirected to the run detail page while the pipeline executes.

Run lifecycle

Status Meaning
queued Accepted, waiting for a worker
running Pipeline in progress
completed Forecast available
failed Error — check run events and notifications

Runs triggered from the UI count against your plan's runs per month. API-triggered runs may incur overage on Growth+ when above quota.

Run detail page

When complete, the run detail page shows:

  • Portfolio metrics — WAPE, bias, series count
  • Forecast chart — historical actuals plus P10 / P50 / P90 bands
  • Series table — per-SKU model, WAPE, trend, demand class
  • Exports — CSV, Excel, PDF (Starter+)
  • Overrides — manual forecast adjustments (Growth+)
  • Events log — pipeline steps and warnings

Compare runs

Use Compare in the sidebar to overlay two completed runs from the same workspace — useful after retraining or data updates.

Scenarios and overrides (Growth+)

  • Scenarios — apply temporary demand adjustments without retraining.
  • Overrides — replace model output for specific series/months; tracked in override impact widgets on the dashboard.

Scheduled runs (Growth+)

Configure automatic retraining on a cadence from workspace settings. Scheduled runs use the same quota as manual runs.

API alternative

Programmatic runs use POST /api/v1/runs with an API key — see API overview.

After the run

  1. Review the Dashboard for portfolio-level KPIs and demand-shift alerts.
  2. Set Notifications so your team is emailed on completion or failure.
  3. Export or schedule reports for stakeholders (Growth+).