In June 2025, Larohra focused on enhancing autoscaling reliability for Azure Service Bus within the azure-sdk-for-net repository. They developed a throttling-based scale-out mechanism in C#, introducing a 2-second delay and a MinimumLastQueueMessageInSecondsThreshold to prevent premature scaling and reduce cost volatility. By updating the scaling logic to require queue-time to exceed a defined threshold, Larohra improved the accuracy of scale-out decisions and lowered operational risk. The work included comprehensive automated tests to validate the new behavior and ensure long-term stability. This backend development effort emphasized monitoring and scalability, resulting in safer, more predictable capacity management for production environments.