The client's dispatch team manually assigned delivery orders to riders across 40 hub locations using a spreadsheet and phone calls, re-balancing load by feel as volume shifted through the day. Peak-hour dispatch decisions were taking dispatchers 12–15 minutes per batch, and there was no visibility into rider capacity across hubs.
An auto-allocation engine that scores open orders against rider location, current load and delivery-window constraints, with a dispatcher override for edge cases. The system integrates with the client's existing rider app rather than replacing it, and includes a live hub-capacity dashboard.
We shadowed dispatchers for a week before writing any code, which surfaced several informal rules (rider familiarity with a route, fragile-goods handling) that weren't in the original spec. The allocation engine was rolled out to one hub first, run in parallel with manual dispatch for two weeks, then extended hub by hub over the following three months.
Average dispatch time per batch dropped from roughly 13 minutes to under 9. Dispatcher headcount needed at peak hours was reduced, with staff reassigned to exception handling rather than manual assignment. The client extended the engagement into an ongoing retainer for feature work.
“Bogo Infotech was the first vendor who pushed back on our original spec — and they were right to. The system they shipped fit how our dispatchers actually work.”
We can walk through what would transfer from this build to yours.