What the AI Tracking Agent does
Proactive “checkpoints” automation:
Wakes up at key milestones (pre-pickup, at pickup, in transit, at delivery, post-delivery).
At each checkpoint it “decides” what’s missing and asks only for what’s needed (empty at prior stop, ETA, tracking enablement, unloaded confirmation, POD, etc.).
Reactive auto-responding (optional):
Reads inbound SMS/email, understands intent (e.g., late-night lumper, breakdown, late ETA vs appointment), replies appropriately, and raises exceptions.
Waits briefly (e.g., 1–2 minutes) so humans can reply first. Backs off if a user is actively chatting.
Uses live data:
Consumes tracking breadcrumbs (driver app, ELDs, link trackers) and computes time-and-distance ETAs and miles-to-next-stop.
Factors appointment windows and stop timing.
Multi-channel messaging:
Texts the driver; emails the dispatcher; presents a single, consolidated chat thread for your team.
Identifies who is speaking (e.g., “sent as Melissa”) and can run under a dedicated, branded phone number with call forwarding.
TMS updates and alerts (configurable):
Sets stop arrivals/departures; creates/resolves on-hand alerts; optionally moves loads to In Transit (not Delivered by default).
Updates a “tracking reference” style field with current status (e.g., “Stop 1 ETA 10/16 14:00 – Chain AI”).
Exceptions (examples):
Late vs appointment; tracking stalled; delivered but not closed; invalid/VoIP phone; fake GPS; permission issues; after-hours lumper; breakdowns.
Customization:
Workspace/pod-level; customer- and carrier-specific conditions; add checkpoint-specific instructions (e.g., require selfie in safety vest, warehouse directions).
Language-aware (e.g., Spanish).
How to enable and configure (recommended rollout)
Start with proactive only:
Go to Automation → Agent, load the Sample Agent.
Set timezone, checkpoint times, and any exclusions (customers, carriers/DOTs).
Keep “auto-parse” on (stop updates and alerts).
Add reactive auto-responding after 1–2 weeks:
Turn on reactive responses; set a short wait window (e.g., 1–2 minutes).
Add customer-specific instructions at checkpoints as needed.
Keep static rules for specialty SOPs:
Retain narrowly defined rules (e.g., warehouse app, safety-vest selfies).
Retire redundant reminder pings replaced by the agent.
Scope and segmentation:
Configure at workspace/pod level; use team members to route notifications.
Use exclusions for regular carriers/projects that don’t need check calls.
Best practices
Make team members explicit on loads so notifications reach the right people (defaults: “email only when offline”).
Avoid running two tracking providers in parallel long-term; it depresses compliance and confuses drivers. If you must, tone down cadence temporarily.
Use the agent for core check calls; use static rules to enforce narrow customer SOPs.
Add multilingual scenarios (e.g., Spanish) where relevant corridors require it.
Common exceptions the Agent handles
Tracking stalled (configurable stall threshold).
Late vs appointment (compares ETA/arrival vs appt; can alert and ask “what happened?”).
Delivered but not closed (can prompt team to close or auto-handle per TMS rules).
After-hours lumper (reassure, collect receipt, escalate).
Permission issues/fake GPS/VoIP (diagnose, instruct fixes, avoid bad data).
Breakdown/inclement weather (acknowledge, collect plan/updated ETA).
FAQs
Does this replace my team? No. It handles routine work, escalates exceptions, and yields when you engage.
Can we customize per customer? Yes—conditions and checkpoint instructions.
Can we limit to certain loads? Yes—by workspace/pod, customer, carrier/DOT, and more.
Will it spam drivers? No—it’s conversation-aware, throttled, and waits for humans.
What’s the driver onboarding? Simple SMS link; app is optional; driver-code autofill supported.
Is it multilingual? Yes (e.g., Spanish).
Which TMS updates are supported? Stop arrivals/departures, on-hand alerts, status notes/fields; moving to In Transit (not Delivered by default). Customer-specific behaviors are configurable.