Forecasting Skill
Forecast future metric values based on historical data. Use this skill when users want to predict future trends, estimate upcoming values, or understand what might happen based on past patterns.
Experimental Feature
This is an experimental feature that is still under active development. Its behavior may change in future releases, or the feature may be removed.
How It Works
The Forecasting skill uses time series analysis and statistical models to predict future metric values based on historical data. When activated, the assistant:
- Uses the Visualization skill to access historical data
- Analyzes historical patterns and trends
- Applies forecasting algorithms to predict future values
- Creates visualizations showing historical data and forecasts
- With data sharing enabled, provides natural language summaries of predictions and trends
The skill helps you understand what might happen in the future based on your historical data, enabling better planning and decision-making.
Examples
Sales forecasting:
User: "What will sales look like next quarter?"
Assistant: [Analyzes historical sales, generates forecast, shows trend visualization]Revenue prediction:
User: "Forecast revenue for the next 6 months"
Assistant: [Uses past revenue data, creates 6-month forecast, explains trends]Growth projection:
User: "Predict customer growth over the next year"
Assistant: [Analyzes growth patterns, projects future values, visualizes forecast]Other use cases:
- “Forecast revenue for the next 6 months”
- “What’s the forecast for website traffic?”
- “Based on past data, what should we expect?”
- “Project future costs based on historical trends”
Limitations
- Requires the Visualization skill as a dependency
- Forecasting accuracy depends on data quality and historical patterns
- Works best with consistent time series data
- Results interpretation requires data sharing to be enabled
- Forecasts are estimates based on historical patterns and may not account for external factors