How would you handle a situation where a deployment failed in production?

If a deployment fails in production, the first step is to roll back to the previous version. Simultaneously, I would initiate a thorough investigation into the root cause of the failure. This includes analyzing logs, monitoring metrics, and collaborating with the development and operations teams to identify and resolve the issue. Implementing automated rollback mechanisms and having a well-defined rollback plan beforehand can expedite the recovery process.