Maritime & Logistics
Global logistics enterprise - email classification and routing
Customer service across 28 agencies globally, with fragmented inboxes and inconsistent regional handling
The challenge
A top shipping and logistics enterprise struggled with fragmented customer service mailboxes across 28 agencies globally. Manual email routing created delays, misrouted inquiries, and inconsistent handling between regions - turning what should be a routine operational function into a source of customer churn and rising support cost.
What we built
We implemented a hybrid AI system with transfer learning and multilingual support that automatically classifies and routes emails within the CRM. The system uses both ML models and LLMs, with human validation for low-confidence cases. CI/CD pipelines retrain the models continuously as language and inquiry patterns shift.
What changed
Classification accuracy across categories
Of global email activity automated
Reduction in email response time
Built with
Azure ML · LLMs · Multi-lingual NLP · Transfer Learning · CI/CD Pipeline
Timeline
Multi-year program (2024–2027)