A system and method for handling one or more dependency services hosted by one or more dependency servers for an upstream service hosted by an administrative server in a distributed computer architecture is provided. The present invention provides for identifying any abnormality in the behavior of the dependency services on the basis of metric values associated with service-parameters of said dependency services. Further, the resiliency services are enabled in order to handle one or more faltering dependency services based on the faulty metric values associated with the service-parameters. Yet further, the one or more faltering dependency services are continuously monitored, and one or more resiliency services are withdrawn once the fault in said dependency services is resolved. Yet further, the present invention provides a conversational bot interface for managing the administrative server and associated dependency services.