How would you optimize a slow-running SQL query that is pulling from a high-volume transactional database?
