Part 7/13:
Moving from traditional data collection—via surveys, manual input—to contemporary digital methods, Harish illustrates how data now flows through sophisticated pipelines. These pipelines often include an enrichment layer, where data is semantically connected to build user profiles or contextual understanding.
For example, combining user activity with environmental data (weather, location) and device info enables richer, more precise inference. He emphasizes enriching data responsibly, connecting data points only when necessary, and ensuring robust denormalization—linking structured, unstructured, and semi-structured data—for better AI performance.