Worked on the apache/spark repository to enhance observability in Spark Structured Streaming by improving the readability of progress metrics. Addressed SPARK-53491 by implementing a targeted fix in Scala that ensures inputRowsPerSecond and processedRowsPerSecond are rendered in plain numeric format rather than exponential notation within the metrics JSON. This adjustment makes the metrics more accessible and user-friendly for operators monitoring streaming jobs. The change was validated against existing metrics pipelines to confirm there was no impact on backward compatibility. The work leveraged skills in Scala, data engineering, and stream processing, focusing on precise, maintainable improvements to the metrics reporting pipeline.