Logistics Agent Zero: Watchdog
Watchdog is Getcho’s real-time delivery monitoring agent — it watches every active delivery, every minute, and alerts your team only when something needs attention. No one should be staring at a tracking link. And certainly, no one should be shuffling between 15 tracking links.
Watchdog is Getcho’s first publicly consumable logistics agent. It’s like PagerDuty meets logistics.
Sources:
Actions:
Watchdog lets logistics managers set parameters for scheduled and ongoing deliveries. Like, for example:
- Driver stop tolerance – the amount of time a driver can stop before Watchdog investigates
- ETA swing tolerance – how much a delivery ETA can change before Watchdog investigates
- Or, just plain ETA cut-off times. Receiving hours are 9:30am – 3:30pm; we have no tolerance for anything later.
- Route deviation – does the driver have other stops? Fleet APIs don’t always expose this information. This is useful for painting a picture.
That last one’s interesting. Some dispatchers see a driver off the beaten path and reflexively assume something’s wrong. But this will look a bit closer, for example, it will:
- run a reverse geocode
- collate with Google Places
- see that our driver’s at a gas station
Why don’t we check back 15 minutes later? Easy. Watchdog runs all of the above checks each minute. No one should be staring at a tracking link.
How it works
Internally Watchdog draws from sources like Radar, Google Street View as well as millions of Getcho delivery data points. Uber’s H3 library along with some custom logic helps detect stops, diversions, and movement. Watchdog checks in each minute, evaluating 13 different types of incidents.
When Watchdog sounds the alarm
Incident detection by itself is a dog with no bite. And Watchdog has bite.
It interfaces directly with Getcho Recovery Workflows. These have the following abilities:
- Reaching out to fleet support and drivers directly
- Pinging stakeholders – ops, the shipper, the recipient – to see if the updated timing still works
- Deassigning drivers for fleets that support it
Getcho data shows that driver check-ins speed up frozen deliveries by 18 minutes on average. And on the flip side, delays don’t necessarily have to be bad. When stakeholders are given the chance to make requisite updates, customer sentiment is overwhelmingly positive. During the recent nor’easter where entire cities shut down transportation, Getcho supported dozens of seamless reschedules.
If you or your dispatchers regularly tab between dozens of tracking links then Watchdog can help.
Stop Staring at Tracking Links
Let Watchdog monitor every delivery, every minute — so your team only gets involved when it matters.
Book Your Free Demo →Frequently Asked Questions
What is Watchdog? Watchdog is Getcho’s real-time delivery monitoring agent. It connects to fleet APIs (Roadie, Uber, DoorDash, Frayt, Bungii) and any public tracking link, checks every active delivery every minute, and alerts your team via email, SMS, WhatsApp, phone, or Slack when something needs attention.
What kinds of issues does Watchdog detect? Watchdog evaluates 13 different incident types including driver stops exceeding tolerance, ETA swings beyond threshold, missed delivery windows, and route deviations. It uses reverse geocoding and Google Places data to add context — so it won’t flag a driver who’s just at a gas station.
Do I need to use Getcho’s delivery network to use Watchdog? No. Watchdog works with any public tracking link in addition to first-party fleet APIs. If you can track it, Watchdog can monitor it.
What happens when Watchdog detects a problem? Watchdog interfaces with Getcho Recovery Workflows — it can reach out to fleet support and drivers directly, ping stakeholders to confirm updated timing, or deassign drivers for fleets that support it. Getcho data shows driver check-ins speed up frozen deliveries by 18 minutes on average.
How is Watchdog different from delivery tracking software? Tracking software shows you where a driver is. Watchdog actively monitors every delivery against your parameters and takes action when something goes wrong — it’s the difference between a dashboard and an on-call ops team.
