Worked on the apache/kafka repository to implement retention-aware restoration for Kafka state stores, focusing on reducing startup latency and unnecessary data processing. Leveraged Java and Kafka Streams to expose retention periods and initiate restoration from relevant changelog timestamps, rather than replaying all historical records. This approach improved cluster availability and resource efficiency by restoring only the necessary data, optimizing the state store management process. The work addressed KAFKA-13499, ensuring traceability through linked commits. Demonstrated expertise in distributed systems and the Consumer API, delivering a targeted feature that enhances the operational performance of Kafka clusters without introducing new bug fixes.