Worked on enhancing reliability for log ingestion in the DataDog/datadog-serverless-functions repository by implementing improved error handling for log intake processes. Developed a backend solution in Python that stores failed log events when intake errors occur, ensuring that data is preserved rather than lost during failures. The approach included updating the DatadogHTTPClient to resolve futures immediately and raise exceptions for failed submissions, which enables faster detection and remediation of ingestion issues. Comprehensive unit tests were added to validate error signaling and failure-path behavior, reflecting a focus on robust backend development, error handling, and test-driven reliability improvements within serverless environments.