Building a Data Platform: Tagging, Streaming, Predictions

#35 - AirAsia Super App Transformation and Lessons from COVID-19 Pandemic - Pablo Sanz

Topic 4 of 15

Published Apr 19, 2021 · Clip 21:37–27:36 (5:59)

Pablo argues that data's value lies in building efficient, reusable user experiences around it rather than siloed use cases. He describes the design system Phoenix and publishing platform COSMOS enabling self-tagging of data, a move from batch ETL to streaming for real-time processing, and the Natch intent prediction platform decoupled from vertical-specific decisions. He also covers the experimentation platform for scaling A/B tests from ~100 to 1,500-2,000 per year, revenue management as a platform, and CleverTap for lifecycle communication.