
Logistics
Your cargo moves fast. Now your answers do too.
AI agents for status questions, freight quotes and documents.
Critical workflows.
Booking, container and arrival status, answered on the spot
Customers ask on WhatsApp or email where their container is. The agent checks the booking against your latest export and replies with status, vessel and ETA. Anything that doesn’t add up goes to your team.
- Booking and container lookup
- Vessel ETA and arrival updates
- Delay explanations from your own data
- Handoff when records don’t match

Freight quotes built from your tariff, never guessed
The agent collects origin, destination, equipment and dates, then prices from your tariff sheet. The math is done by code, not the model, and any rate outside policy waits for your approval.
- Collects route, equipment and cargo details
- Prices from your current tariff
- Flags special cargo and surcharges
- Approval before committing a rate

Arrival, release and missing-document notices that go out on time
The agent follows each shipment’s milestones and writes to the consignee before they have to ask: arrival notice, cargo release and which documents are still missing to clear it.
- Arrival notices to consignees
- Release notices once requirements are met
- Missing-document reminders
- Follow-up until the file is complete

Bills of lading and invoices, read and loaded without retyping
Bills of lading, commercial invoices and packing lists arrive in every format. The agent extracts each field, checks it against the booking and sends anything doubtful to a person.
- Bill of lading data extraction
- Cross-checks against booking and manifest
- Confidence shown for every field
- Loads into your system or spreadsheet


A shipping line that stops answering “where is my container?” by hand
A mid-sized shipping line serves importers and forwarders on two routes. Its customer service team spends most of the day answering status questions on WhatsApp and email, chasing missing documents and preparing freight quotes from a shared tariff spreadsheet.
- A workshop with the commercial and operations managers picks one workflow to start: booking and container status.
- The team shares the daily exports it already produces: bookings, containers, vessels and estimated arrivals. No IT project needed.
- The agent answers on WhatsApp and email, checking each booking against the latest export and handing off any case where the data doesn’t match.
- Arrival and release notices are added next, written from the same data and reviewed by the team during the first weeks.
- Freight quotes follow, priced by code from the tariff sheet, with every rate outside policy sent for approval.
What changes
Status questions get an answer at any hour, consignees hear about arrivals and missing documents before they call, and the customer service team spends its time on exceptions and on customers who need judgment. Every conversation and action is on the record for the managers to review.
Example scenario with Krill Austral Shipping, a fictional company. It shows how an agent would work, not a customer result.
Why Baguals.
A future-ready operation, not a forced one
You hand us the process and we do the rest. We prepare the operation underneath: the data, the rules and the handoffs, so agents can take the repetitive load.
Starts with the spreadsheets you already have
The pilot runs on the booking, container and tariff exports you produce today. A direct connection to your shipping system comes later, when the workflow has earned it.
Rates and commitments stay in your hands
Prices come from your tariff through code, never from the model. Committing a rate or changing customer data needs human approval by default.
Every answer on the record
Each conversation and action lands in a log that can’t be edited, so your managers can see what the agent said, which data it used and who approved what.
Solutions for this industry
Questions, answered.
Do we need our IT team to get started?
No. The pilot runs on the spreadsheets and exports you already have, such as booking lists, container status reports and tariff sheets. You refresh them on a schedule and the agent reads the latest version. A direct connection to your shipping system can come later, once the workflow has proven itself in production.
Can the agent commit to a freight rate?
Only within the rules you set. Prices are calculated by code from your tariff, never invented by the model. Committing a rate is treated as a sensitive action, so by default a person on your team approves it before the quote reaches the customer. You decide which cases, if any, can go out without review.
What happens when the data doesn’t match?
The agent says so instead of guessing. If a container number isn’t in your records, a booking shows conflicting dates or a bill of lading field is unreadable, the case goes to a person with the context already gathered. Every step is logged, so your team sees exactly what the agent saw and did.
Which channels can the agent work on?
WhatsApp, email and web chat are the usual starting point for shipping lines and forwarders, because that is where customers already ask. Voice, Slack and Teams can be added as the operation needs them. The agent keeps the same rules, the same data and the same record across every channel.
How do we start?
With a workshop with your leadership to choose one workflow, usually shipment status, freight quotes or arrival and release notices. We build it on your own data, agree on a metric upfront and get it running in weeks. Once it works, we move on to the next workflow.

