"Where is my shipment?"
The single highest-volume enquiry in the industry, answered manually hundreds of times a week from data that is already in a system.
Home / Industries / Logistics
AI Automation for LogisticsAI Customer Operations for Logistics
Stop paying skilled operations people to answer "where is my shipment?" — and stop retyping delivery notes into your TMS by hand.
The problem
Your operations team spends its day on status updates
In freight forwarding, haulage, 3PL and last-mile, the same two jobs consume the team: telling people where things are, and retyping documents between systems.
Paper everywhere
Delivery notes, PODs, CMRs, customs paperwork and invoices arriving as scans and photos, then keyed in twice.
Exceptions found late
A delay nobody flagged becomes an angry customer call instead of a proactive notification.
Booking by email
Collection requests arriving as free text in a shared mailbox, transcribed by hand into the TMS.
Out-of-hours silence
Nights and weekends are when freight moves and when nobody is available to answer the phone.
Reporting by spreadsheet
Customer service levels assembled by hand at month end, too late to act on.
The solution
The status desk and the document desk, automated
Two systems that share one integration into your TMS: one answers customers across every channel, the other turns paperwork into structured records.
- Live shipment status from your TMS, on chat, email and phone
- PODs, delivery notes and invoices read, validated and filed
- Collection requests parsed from email into TMS bookings
- Proactive delay notifications before the customer chases
- Anything commercial, claims-related or disputed goes to a person
Where the line is drawn
agreed before go-live✕ Negotiate rates or accept a claimtransfers
✕ Give customs or compliance rulingstransfers
✕ Commit to a delivery date it cannot verifytransfers
✓ Status, ETA, POD retrieval, booking, documentshandles
Business outcomes
What changes in the operations room
Self-service statusRoutine tracking enquiries answered without a person
No double keyingDocuments read once and written straight into the TMS
Earlier exceptionsDelays flagged before the customer notices
Covered overnightEnquiries answered when freight is actually moving
We publish measured percentages only from live operations that have agreed to share their numbers. Your baseline is measured during the pilot.
Estimate
What does the status desk cost you?
This one measures time, not lost revenue — because in logistics the cost is your team's hours.
Phone, email and portal messages asking where something is.
Count the context switch, not just the conversation.
Salary plus employer costs, divided by working hours.
Estimated monthly cost of this work
$0
spent answering routine status enquiries
Enquiries per month0
Hours consumed per month0
Estimated annual cost$0
Calculation: weekly enquiries × 4.33 weeks × handling minutes ÷ 60 × your hourly cost. This is the cost of the work today, not a saving we promise — no automation removes 100% of it, and the realistic share is measured during your pilot.
Demo
The two desks, running
app.workgenesis.ai/logistics
LIVE
Customer enquiries
Live from the TMS| Enquiry | Channel | Reference | Answered from | Outcome |
|---|---|---|---|---|
| Where is my consignment? | CON-88214 | TMS tracking | Resolved by AI | |
| Has it been delivered? | Portal chat | CON-88190 | POD on file | POD sent |
| Can I get a copy of the POD? | Phone | CON-87766 | Document store | Emailed |
| Book a collection Thursday | New | Parsed to booking | Draft in TMS | |
| Damage on delivery — claim | CON-88103 | — | Escalated to claims | |
| Renegotiate our rates | Phone | Account | — | Escalated to sales |
Document processing
Extract → Validate → Post| Document | Type | Matched to | Status | Destination |
|---|---|---|---|---|
| POD_88190_signed.jpg | Proof of delivery | CON-88190 | Validated | TMS + doc store |
| CMR_44127.pdf | CMR note | CON-88214 | Validated | TMS |
| Carrier_invoice_9921.pdf | Purchase invoice | PO-44120 | Rate mismatch | Finance review |
| Delivery_notes_batch.zip | Scanned batch | 184 consignments | Validated | TMS |
| POD_photo_blurred.jpg | Proof of delivery | Unreadable | Low confidence | Human queue |
Exceptions detected before the customer called
Proactive| Consignment | Issue | Detected | Action taken |
|---|---|---|---|
| CON-88231 | ETA slipped past agreed window | Automatic | Customer notified with revised ETA |
| CON-88245 | No scan event for 14 hours | Automatic | Flagged to operations |
| CON-88250 | Delivered, POD missing | Automatic | Driver chased for upload |
| CON-88262 | Address unresolvable | Automatic | Customer asked to confirm |
Layout example showing the exception queue, not client data.
Use cases
What it handles day to day
Track & trace
Live status and ETA pulled from your TMS on whichever channel the customer used.
POD retrieval
Proof of delivery found and sent without an operator opening the document store.
Collection booking
Free-text email requests parsed into structured TMS bookings for review.
Document processing
PODs, CMRs, delivery notes and carrier invoices extracted and validated.
Invoice matching
Carrier invoices checked against agreed rates, mismatches sent to finance.
Exception alerts
Missing scans, slipped ETAs and failed deliveries surfaced automatically.
Proactive notification
Customers told about a delay before they ring to ask about it.
Service reporting
On-time performance and enquiry volumes reported without a spreadsheet.
Integrations
It reads from the systems you already run
Everything above depends on one thing: getting reliable data out of your TMS. That is the first thing we verify, before anything is quoted.
TMSTransport management
WMSWarehouse system
EDIEDI feeds
APICarrier APIs
OLShared mailboxes
ERPFinance / ERP
If your TMS has no API and no usable export, we will tell you that in discovery rather than build something fragile on top of screen scraping.
How it works
Status desk first, documents second
Step 01
Discover
We measure your enquiry mix and confirm what your TMS can actually expose.
Step 02
Design
We define which enquiry types the AI may close and which always reach a person.
Step 03
Deploy
Status desk goes live first on one channel, supervised, so accuracy is proven on real traffic.
Step 04
Improve
Once status is stable we add document processing on the same integration, then report monthly.
Expand your AI automation
What operations teams add next
Step 2
Document operations
PODs, CMRs and carrier invoices processed and matched automatically.
Step 3Voice for the status line
The same answers on the phone, including overnight cover.
Step 4Customer portal chat
Self-service tracking embedded where your customers already log in.
Step 5Quote & tender response
Rate requests captured, structured and followed up consistently.
FAQ
Questions operations directors ask
Often, through an API, a read-only database connection, an EDI feed or a scheduled export. We verify this in discovery and put the method in the proposal. If there is genuinely no reliable way to read your data, we tell you before you spend anything.
It only reports what your TMS says, with the timestamp of the last scan event, and it never estimates a date the system has not produced. If the data is stale or missing, it says so and escalates rather than guessing.
No. Damage, loss, claims and rate negotiations are escalated to a named person with the full history attached. Those conversations carry commercial and legal consequences and should not be automated.
That is your decision, not ours. Most operations we speak to are short-staffed rather than over-staffed, and use the recovered hours for exception management and customer relationships. We will not build you a business case that assumes redundancies you have not decided on.
Start with one week of your enquiry log
Send us a week of tracking enquiries and a sample of your PODs. We will show you what share is genuinely automatable — including the part that is not.