AI service
Amazon Forecast
Managed time-series forecasting.
Key points
- Provides managed time-series forecasting for future numeric values.
- Uses historical target time series and can include related time series or item metadata.
- Supports scenarios such as demand, inventory, staffing, traffic, or revenue forecasting.
- Can generate forecasts without teams building forecasting algorithms from scratch.
- AWS closed Amazon Forecast to new customers on July 29, 2024, while existing customers can continue using it.
When to use it
- Choose Forecast in legacy exam-style scenarios that ask for managed demand or time-series forecasting.
- Use it conceptually for predicting future sales, call volume, inventory needs, or resource demand.
- Use SageMaker or other current forecasting approaches for new custom forecasting workloads.
Exam tips
- Forecast predicts future numeric time-series values; Personalize recommends items to users.
- Lookout for Metrics detects anomalies in metrics, while Forecast predicts future values.
- Because new customer access is closed, keep recommendations for new workloads conservative.
- Forecasting quality depends on historical data quality, seasonality, and relevant related variables.