Worked on backend reliability for the marqo-ai/marqo repository by addressing a StatsD emission issue related to HTTP status code handling. Implemented a fix in Python to ensure all HTTP status codes are cast to integers before being emitted, preventing errors and inaccurate logging caused by non-integer representations. Developed comprehensive unit tests to validate the solution across a range of status codes, improving the reliability of monitoring and telemetry. Focused on backend development, monitoring, and testing, the work enhanced the accuracy of metrics and observability for the system. The contribution centered on robust type handling and thorough validation practices.