To efficiently process Kafka event streams without saving them, first, understand the data flow using Kafka's consumer groups. Then, use in-memory data structures to store relevant information temporarily. Finally, design a REST API interface to deliver the processed data to callers. Familiarize yourself with Kafka's consumer APIs and explore libraries for in-memory data handling.