AI Features

What kind of predictions does the system provide?

The system can generate forecasts for your Primary Metrics based on historical data patterns. As we train machine learning models to provide those predictions, you can expect additional metadata which can help you trust the forecast:

• Information about the model used

• Error metrics to help you assess reliability

• Drift charts to monitor prediction accuracy over time

How can I understand how the system reached its conclusions?

Every answer provided by the system includes an explanation section with:

• A summary of the logic used

• Details of any calculations performed (such as SQL queries)

• Metadata about precision and models used

• The data products accessed to generate the answer

How accurate are the predictions provided by the system?

The prediction accuracy varies based on the quality and quantity of historical data available, as well as the inherent predictability of the metric. Each prediction includes error metrics and a drift chart to help you assess its reliability. The system continuously learns from new data, improving prediction accuracy over time.