How to Use Dynamic Automation Tools to Improve Batch Processing - Page 3

4. Actively hunt down manual processes to automate

Manual processes are slow, error-prone and can sabotage compliance efforts. Eliminate them with automation. The manual processes to be targeted should not just include those that have a time- or date-based component, but should include any process that requires intervention-even processes that potentially fall under a "workflow" designation.

Organizations can survey users for the processes they least enjoy performing. For example, an existing manual process consisting of several inter-related but disparate steps may be viewed as impossible to automate. Workload automation provides wider coverage so each step of the manual process can be transformed into a "job" which can be strung together into a unified, automated business process.

5. Look for events to drive dynamic processing

One critical capability of a workload automation tool is the ability to pay attention to all events that occur within the organization. Actionable events are more than just the completion of a prior job. Possible event scenarios could include a change to specific data in a database, the arrival or disappearance of a file on a file system, a Web service invocation, receipt of a Java Message Service (JMS) message, or exceeding a critical operating system parameter (such as CPU usage).

The ability to react to events is a strong tool for the automation of manual processes because the same cues that a user might look for to determine if or when they should manually perform a process can be detected and used by the workload automation tool to automate the same process. This detection of events is typically much quicker and less error-prone than a human can ever be.

Events expand the scope of automation well beyond the "scheduling" time-based mindset. A centralized workload automation tool can detect and take advantage of events that occur across a variety of disparate servers and applications to drive and control automated processing. Events, therefore, become a kind of glue that can connect disparate-and previously thought of as unconnected-processes together into a more cohesive, automated whole.