A dispatcher's screen that shows what needs attention, not everything

The problem

Dispatchers found problems by scrolling. A case that had gone stale, a route that had failed to build overnight, a warranty waiting on review, a repair that had stalled: each lived in a different list, and the person who noticed was usually the driver.

What we built

One screen per territory. A KPI rail shows open cases, dispatched today, backlog, response time, first-time fix, and techs on shift. Attention lists surface stale cases, routing failures, warranties awaiting review, and jobs stalled in repair. Workload and dispatch tabs show the week by driver, with month-over-month charts behind them. A driver day view, filtered to one technician and today, mirrors the overnight route post with an open-in-maps link.

The one thing we got wrong: it auto-refreshed every five minutes, technicians kept it open during queue work, and the open tabs saturated the org's limit on concurrent long-running requests. We removed the polling the same day and left a manual refresh. The lesson is in how we work: ship the monitor with the feature, so the feature tells you when it's the problem.