Kavpreet Grewal enhanced the apache/spark repository by delivering two Kafka-related features focused on documentation clarity and runtime resilience. Kavpreet developed data visualization diagrams to illustrate Kafka timestamp offset options, improving user understanding and validating the changes through a successful documentation build. Additionally, Kavpreet implemented retry logic for Kafka Admin client calls within the KafkaOffsetReaderAdmin class, generalizing the withRetries helper to support generic types and adding comprehensive unit tests to ensure reliability. Working primarily with Scala and Markdown, Kavpreet’s contributions demonstrated depth in both technical writing and robust engineering, with all new and existing tests passing after local validation.