← Back to the library
C.H. Robinson turns emailed tenders into shipment orders
Added 2026-10-14 · Reviewed 2026-10-14 · C.H. Robinson
Read the source ↓What the company did
C.H. Robinson built an AI Orders Agent that reads customer tender emails and their attachments, interprets the shipment information and creates orders. It handles intake work that otherwise requires someone to read the tender and enter the details.
How it went
In its June 2025 report, C.H. Robinson said the agent built orders in about 90 seconds, processed 5,500 truckload orders per day and saved 600 hours of labor daily. The report does not disclose order accuracy, exception rates or how the labor savings were calculated. Freight Fieldnotes has not independently verified this account.
Worth knowing
- It uses a proprietary system. The company describes freight-specific models and shipment data supporting its broader AI deployment. It does not publish a setup or prompt for reproducing the Orders Agent.
- Speed leaves an unanswered question. The source does not explain how incomplete, conflicting or unreadable tenders are handled.
- For your desk: If you evaluate similar intake automation, check whether the order matches the tender and whether missing information is flagged before the order proceeds.
Sources & provenance
C.H. Robinson Scales Fleet of AI Agents Past 30 ↗
C.H. Robinson · Published 2025-06-12 · Reviewed 2026-10-14
“Reads and interprets emailed tenders, including attachments, and builds orders in about 90 seconds”
Supporting evidence and review
Human review required
- Treat publisher statements as attributed reports, not independent measurements.
- Verify current product availability and access requirements with the provider.
Evidence & limitations
- No public accuracy benchmark or exception-handling specification.
- Model, prompts, integration interface and error rates are undisclosed.
- Reported throughput is not independently verified and is not a prediction for another team.