Worked on the confluentinc/schema-registry repository to improve system reliability by addressing a critical shutdown sequence issue. Focused on backend development and system design using Java, the work involved delivering a targeted bug fix that prioritized shutting down the Leader Elector before kafkaStore and metadataEncoder. This approach was designed to prevent race conditions during shutdown, thereby reducing restart-related risks and supporting smoother production deployments. The change was implemented as a single, well-scoped commit with clear traceability to both a JIRA ticket and a pull request, enhancing maintainability and making the review process more straightforward for future contributors.