Data Normalization

Multi-Provider Data Normalization

Definition

The process of standardizing data formats, units, and schemas across multiple heterogeneous data sources so they can be queried and analyzed uniformly by an AI agent without provider-specific handling.

Examples in the Wild

  • Example 1:Converting sleep metrics from Whoop, Garmin, and Withings into unified schema
  • Example 2:Standardizing heart rate variability (HRV) measurements across providers
  • Example 3:Normalizing GPS and power data from different cycling platforms