ZeroGrad
Capabilities · Forecasting

Weekly Demand Forecast

Healthy
Zero-shot forecastingEndpoint weekly-demandProduction v4Challenger v5
WAPE
12.8%
Bias
1.4%
Interval coverage
91%
Last backtest
2h ago
Production
5 weeks ago
v4
WAPE
12.8%
P95
380ms
Cost / 1K
$0.40
Serving 100% of traffic
Challenger
2 hours ago
v5
WAPE
11.9%
P95
410ms
Cost / 1K
$0.42
Shadow evaluation complete · Not serving traffic

Call it from your product

Sandbox
forecast = ai.forecast(
    capability="weekly-demand",
    series=data,
    horizon=12
)
Response
{
    "prediction_id": "pred_01J9X4S9VD",
    "forecast": [412, 398, 430, 455, ...],
    "p10": [361, 342, 371, 390, ...],
    "p90": [468, 459, 494, 523, ...],
    "version": "v4"
}

Next step

v5 beats v4 on WAPE (12.8% → 11.9%) in offline evaluation. Shadow results confirm it on live traffic.

Activity

Today
  1. 12:00Backtest completed — v5 improved WAPE 12.8% → 11.9%
  2. 10:00v5 created — intermittent series routed to statistical baseline
Yesterday
  1. 06:20Late data from 2 stores — forecasts used last known values
5 weeks ago
  1. 11:00v4 promoted to production

Use it from an agent

Every capability is also a tool. Give your agents calibrated predictions instead of letting them guess.

Private preview

One remote MCP server for the whole workspace. Works with Claude, Cursor, VS Code and any MCP client.

Claude Code
claude mcp add --transport http zerograd https://mcp.zerograd.dev/v1 \
  --header "Authorization: Bearer $ZEROGRAD_API_KEY"
mcp.json · Claude Desktop, Cursor, VS Code
{
  "mcpServers": {
    "zerograd": {
      "type": "http",
      "url": "https://mcp.zerograd.dev/v1",
      "headers": { "Authorization": "Bearer ${ZEROGRAD_API_KEY}" }
    }
  }
}
Tools exposed
  • weekly_demand_forecastForecasting
    Forecast weekly units sold for a SKU × store series, with P10/P90 intervals.
  • get_capability
    Health, production version and live quality metrics.
  • list_capabilities
    Every capability in the workspace.
  • record_outcome
    Send the real outcome for a prediction_id once it is known.

Tools are scoped by API key. Calls are logged, metered and show up in this capability's observability like any other request.