During the month, contributed a targeted fix to the apache/druid repository, focusing on improving Kafka lag calculation consistency within backend data pipelines. Addressed an issue where negative Kafka partition lag appeared by implementing a coherent offset snapshot mechanism, basing lag calculations on the latest and highest ingested offsets. This technical approach stabilized ingestion monitoring and reduced false alerts, enhancing overall observability for stream processing workflows. The work involved updating both tests and documentation to reflect the new offset snapshot logic, as well as refining code by removing brittle references. Utilized Java, Kafka, and backend development skills throughout the process.