Case Study
Cargo Readiness Notification Platform — CMA CGM
Built an enterprise internal platform for CMA CGM Canada to automate cargo release notifications across rail and port operations — replacing a fully manual workflow that caused delays, increased terminal cost exposure, and created communication errors at scale.
The Problem
Manual workflows at logistics scale
Customer service teams validated payments, checked pickup numbers from rail vendors, and emailed customers individually — one container at a time. With hundreds of daily cargo releases across CN Rail, CP Rail, and port operations, this process caused pickup delays, increased demurrage costs, and left teams with no real-time visibility into release readiness status.
Architecture Decisions
How I solved it
Custom data grid over off-the-shelf alternatives
Operational workflows required row-level updates, inline status visualization, dialog-based detail views, and real-time validation — generic grid libraries couldn't support this interaction model cleanly without fighting their defaults.
Debounced autocomplete search architecture
Datasets include hundreds of active containers at any time. Triggering API calls on every keystroke caused visible lag and wasted server resources. Debouncing reduced query volume by ~70% while keeping search snappy.
Export-first operational design
In logistics, audit trails are critical. Teams needed to generate Excel reports of release history, filtered datasets, and notification logs for internal review. Designed export as a first-class feature, not an afterthought.
Standardized service and error patterns
The platform integrates with CN Rail (CNRR), CP Rail (CPRR), finance validation, and customer notification APIs — each with different response shapes and failure modes. A shared service abstraction kept error handling consistent and made adding new integrations straightforward.
How It Worked
Automated release lifecycle
Once a container was flagged as ready, the platform ran a validation chain: Bill of Lading verification → payment clearance → customer credit check → invoice validation. Only on full pass did the system retrieve the pickup number from the rail vendor and automatically dispatch the customer notification. Background jobs polled vendor APIs hourly to sync release status without manual intervention.
Stack
Outcome
−20%
Communication error rate
Hourly
Automated release sync
2 flows
Rail + port unified