Grab's Customer Data Platform introduced Scenarios, a self-serve feature enabling real-time personalization at scale. The system combines event triggers, geo-fencing, historical user data, and predictive models to deliver dynamic experiences like mall offers and traveler recommendations. Built on Apache Flink with integration to Kafka and DynamoDB, Scenarios achieved over 3% conversion uplift in production use cases. The platform eliminates engineering bottlenecks by providing marketers with a GUI to configure complex event sequences, validate with synthetic data, and deploy within an hour. Over 12 production implementations demonstrate measurable business impact across Southeast Asia.
Table of contents
IntroductionAbout Grab’s CDPThe need for real-time personalizationProblem statementCDP Scenarios: Real-time personalization made simpleProven impact: Real-world successTechnical architecture: Low latency, high reliabilityAdvancing the future of real-time personalizationConclusion: Revolutionizing real-time personalizationJoin usSort: