Voice AI for MercuryGate

An AI voice agent foryour MercuryGate TMS.

Multi-mode operations produce multi-mode phone calls. The agent answers them, identifies the shipment, and posts the outcome back through the same API and EDI layer your other integrations already use.

REST API and EDI Multi-mode aware Scoped credentials
Live call
Answered in 1 ring ยท 0:00
Live
Calling about shipment 55710, the LTL pickup that was supposed to happen today.
I see it. Pickup was scheduled for this morning with a noon cutoff. What happened?
Driver got there and the freight was not ready. We are rescheduling for tomorrow.
Logging a failed pickup with that reason and moving the pickup to tomorrow. Your rep is being notified.
Posted to MercuryGate
Shipment55710EventFailed pickupReasonFreight not readyRescheduleNext business day โœ“

What it covers

ShipmentsStops & appointmentsCarriersCustomersTruckloadLTLIntermodalParcelStatus events
24/7
Answers and dials, nights and weekends included
<2 rings
Average time to pick up an inbound call
Live
Reads and writes your TMS during the call
EN / ES
English and Spanish, natively

Every mode has its own phone call. They all land on the same team.

Running truckload, LTL, intermodal, and parcel on one platform is the point of MercuryGate. The phone does not respect that tidiness. A missed LTL pickup, a truckload driver at a closed dock, and a customer asking about an intermodal container all hit the same team, each needing a different answer from a different part of the record. Triage plus live lookup is exactly what a well-built voice agent does better than a person juggling three screens.

What it handles

Triage across every mode you run

Mode-aware call handling

Identifies whether the call is about a truckload, an LTL shipment, or an intermodal move and follows the right script for each.

Check calls and status updates

Captures position, loaded status, and ETA on the phone and posts the update against the shipment.

Failed pickups and exceptions

Structured intake with the reason code your team actually needs, logged and escalated by your rules.

Customer status calls

Answers shipper questions from the live shipment instead of taking a message for an account manager.

Appointment confirmations

Calls facilities to confirm or reschedule dock times and updates the stop when the answer changes.

Document follow-up

Chases the paperwork holding up settlement and confirms it against the shipment when it arrives.

MercuryGate integration

The same connection layer your ERP already uses

MercuryGate is built to connect: REST APIs, traditional EDI, file-based integration, and pre-built connectors into ERP systems. A voice agent is just another integration on that layer, with one difference. It produces data from a phone conversation, which is the one channel your EDI feeds have never been able to cover.

  • Reads shipments, stops, appointments, carriers, and customers in real time
  • Posts status events, exception reasons, and notes back to the shipment
  • Uses REST where available and your existing EDI flows where that is the better path
  • Sits alongside your ERP connectors without touching them
  • Transcript and recording attached to every update it writes

What we scope with you

Which modes generate the most phone traffic for your team
Which shipment objects the agent may read and write
Reason codes and exception structures your team already uses
Per-customer routing and escalation rules
Whether inbound triage or outbound check calls go live first

Have a mode-specific workflow in mind?

How it works

One mode at a time

01

Connect

Scoped credentials, read-only first, so you can hear real calls against real shipments before anything is written.

02

Model the call types

We script per mode, because an LTL pickup failure and a truckload detention call are not the same conversation.

03

Go live on the noisiest mode

Usually LTL pickups or truckload check calls. One workflow, proven, before the next.

04

Expand

Add modes and call types from what you hear in the transcripts, not from a rollout plan written in advance.

Questions MercuryGate users ask us

Yes, and it has to, because the questions are different. It identifies the shipment first, sees the mode on the record, and follows the script you defined for that mode. Getting this wrong is one of the fastest ways for a generic voice bot to embarrass you, which is why we build it per mode rather than one size fits all.

Whichever is the right path for each workflow, and we tell you which is which. Status updates often flow best through the path your organisation already trusts, while live lookups during a call need the API.

No. It writes shipment-level data through supported paths, the same as any other integration, and it does not touch your ERP connectors.

Yes, and facility calls are often the most valuable to automate because they carry appointment and dock information that otherwise arrives as a voicemail nobody actions until the truck is already late.

We map the agent to the reason codes your team already uses, so the exception it logs is queryable and reportable like every other exception. Free-text notes are the fallback, not the default.

About a week for the first workflow once credentials are sorted, then a couple of weeks per additional mode. Enterprise change control is usually the longer pole, not the build.

Pick your noisiest mode and start there

Book 15 minutes. Bring the call type your team is most tired of and we will show you the conversation, the write, and where it lands in MercuryGate.

Part of our voice AI for freight work, alongside AI voice agents for every other kind of business.