What should we give an operations analyst first?
Start with a specific decision, the process involved and access to people who understand the work. Provide the current definitions and known system changes. A request to build a dashboard is less useful than a clear question such as why one category of cases repeatedly misses its target.
Can they work across billing, sales and service operations?
They can, but define the priority domain and required knowledge. Billing reconciliation, lead routing and service queues have different records and controls. A broad remit needs realistic capacity and access to the relevant functional owners.
Do we need process-mining software?
No. Many investigations can begin with checked exports, queries and case walkthroughs. Process-mining tools become useful when suitable event data and a clear question justify them. Buying a tool does not establish that the recorded events accurately represent the work.
Can they automate the process they analyse?
They may support requirements, testing and simple approved changes where their skills match the task. Production automation needs the appropriate technical ownership, access controls and maintenance plan. Separate identifying a repetitive task from deciding it is safe and worthwhile to automate.
How do we know an improvement worked?
Agree the intended outcome and baseline before the change. Check quality and downstream effects as well as speed, and record other factors that changed during the trial. A measured improvement is more useful when the team can explain the conditions under which it occurred.