AI agents for logistics

Mirage Metrics builds AI agents for freight forwarders and fleet operators: quote requests and freight documents checked into your TMS, trucks and fuel checked against the plan. Only the exceptions reach your team.

Book a call

Quote requests and freight documents arrive faster than a desk can key them.

Every request priced by hand, every freight document re-keyed into the TMS. Mirage reads each one, checks it and drafts the answer. A person only sees what does not match.

  1. Read
  2. Check
  3. Draft
  4. Route

The quote or the TMS entry, written through the TMS's own API. No new software layer for the team.

Inbox
As received

Hello, could you quote FTL Barcelona to Lyon, tautliner 13.6 m, 33 pallets, 24 t, loading next Tuesday? Thanks.

Checked against Contracted rates
Lane
Barcelona to Lyon
Match
Equipment
Tautliner 13.6 m
Match
Load
33 pallets, 24 t
Match

Quote drafted at the contracted rate.

Illustrative example: the documents and values are not a client's. The checks are the ones described above.
Container terminal in evening light: gantry cranes over a container ship and rows of stacked containers

In production at OFRET Groupe and Transwin

A freight forwarder running two entities on one TMS, and a logistics operator built around port flows, with 550 specialized trucks.

90%+

Of standard RFQs answered automatically

OFRET Groupe

70%

Less manual data entry on freight documents

Transwin

12%

Lower fleet diesel consumption

Transwin

5

Systems live across freight, fuel, mining and fleet

Transwin

“We started with one workflow, and Mirage has been alongside us at every step since. Every process we brought them, they built and put into production. Today they are not a supplier we buy software from. They are our AI partner.”

Yassine Bennis, CEO, Transwin
Before
OFRET for full truckload, SWILL for groupage: one TMS, two pricing logics, two carrier networks, and every quote request priced by hand.
Deployment
A Quoting Agent in two separate environments, one per entity, built inside the group's Logitude World TMS through its API. Document automation first, then RFQ automation, then a verification and exception layer.
Now
Neither team re-keys documents or prices standard RFQs by hand any more. Anything outside the standard pattern goes to an operator before it goes out.

Also in production at

Does it work with my existing stack?

Your TMS stays the system of record. In production on Logitude World at OFRET Groupe and SAP Business One at Biodéal; CargoWise through its eAdaptor interface; telematics, fuel terminals and tank probes on the fleet side. No system replaced.

  • Logitude World
  • CargoWise
  • Descartes
  • Magaya
  • Freightview
  • BluJay
  • SAP
  • Oracle
  • MicrosoftDynamics
  • Salesforce

Where a documented interface exists, Mirage connects to it. Where none exists, our engineers build the connector during deployment.

How a deployment runs

  1. Observe
  2. Map
  3. Build
  4. Run
  5. Expand

Most deployments reach production in 4 to 6 weeks, depending on the number of document types, workflows and systems involved.

See how we deploy

Security and governance

  • Deploy anywhere
  • Control every action
  • Prove what happened

Access permissions, validation workflows and audit logs remain fully controlled by your organization. Signed NDA on request.

See security and data handling
Ship-to-shore cranes and stacked containers along a container terminal, seen across the water

For freight desks and fleet operations alike

Mirage Metrics builds and operates AI agents for freight forwarders and logistics operators. Two kinds of team on the same platform: what changes is where the time goes.

Frequently asked questions

  • AI document processing for freight is software that reads shipping documents (bills of lading, CMRs, packing lists, customs declarations, invoices), extracts the required data fields, validates them against shipment records, and sends clean structured data directly into your TMS, ERP, or customs system without manual entry.

  • OCR reads characters on a page but produces raw text with no structure or meaning. Mirage understands what a bill of lading is, which fields it contains, and how those fields map to your systems. It works without templates, handles documents from any carrier in any format, and flags discrepancies rather than passing bad data downstream.

  • Mirage processes bills of lading, CMRs, airway bills, sea waybills, packing lists, commercial invoices, customs declarations, certificates of origin, and dangerous goods declarations. New document types are added through model training rather than template configuration.

  • Mirage connects to the TMS, ERP, WMS and customs platforms a client already runs, including CargoWise, SAP and Oracle, through their APIs or a file exchange; when a system has no API, Mirage builds the connector as part of the deployment. Data flows from document to system without intermediate manual steps and without changing your existing stack.

  • Mirage publishes its results per client, on the case-study pages. At Transwin, freight document processing runs with 70% less manual data entry. At OFRET Groupe, more than 90% of standard freight quote requests are answered automatically. Mirage publishes no accuracy percentage: the share of documents that need a human check is measured on your own documents during the proof of concept.

  • Mirage flags the specific field, shows the operator exactly what was found and why it was uncertain, and routes the exception for review. The operator corrects that field rather than processing the full document from scratch, and each resolution improves the handling of future documents of the same type.

  • Mirage is built for two kinds of logistics team, at equal weight: freight forwarders, customs brokers and supply chain teams that process high volumes of shipping documents daily, and carriers and fleet operators running trucks, fuel and machines across sites. It is particularly effective for operations handling documents from multiple carriers, origins, and formats where template-based OCR systems require constant maintenance.

  • Deployments reach production in 4 to 6 weeks, depending on the number of document types, workflows, and systems involved. A proof of concept on your own documents comes first, one week after the first meeting.

  • Mirage integrates into your existing operational environment without requiring system replacement. Access permissions, validation workflows, and audit logs remain fully controlled by your organization. Every extraction, validation, and action performed by the agent is traceable and can be reviewed at any time.

Bring us the operation you need to improve.

Mirage will map the workflow, identify what can be automated and define the fastest path to production.

Start with a written plan

Tell us where your team loses time. We send back which workflows are worth automating, what a deployment involves on your systems and the shortest path to production. No commitment.

We use your email to reply to you and, now and then, to tell you about Mirage. You can opt out at any time. Privacy policy

Pick a time that works.

30 minutes with a Mirage engineer.