AI-Powered Logistics Platform
Intelligent package tracking and customer service powered by conversational AI, real-time data, and workflow automation.
The Challenge
A Caribbean shipping company handled customer inquiries entirely by email and phone. Tracking a package meant emailing the office and waiting for a response. Customers in different time zones often waited hours or days for simple status updates. The staff spent most of their day answering the same questions: Where's my package? What are your rates? What can I ship?
Our Solution
We built a four-component AI system that handles customer queries in real time. An AI chatbot on the company website classifies every question by intent — tracking, history, billing, or general — and routes it to the right backend. Package tracking queries hit a Firebase Cloud Functions API for live data. General questions pull from a curated knowledge base. Everything runs on automated workflows with zero manual intervention, and customers get answers in seconds instead of hours.
AI Chatbot
A customer-facing conversational AI that handles tracking queries, shipment history, billing questions, and general business FAQs — instantly, 24/7.
Intent Classification
AI classifies every customer message into the right category — track, history, billing, or general question — and routes it to the appropriate backend system.
Real-Time Package Tracking
Live package status via Firebase Cloud Functions. Customers get instant updates on their shipments without waiting for email replies.
Business Knowledge Base
A curated knowledge system that the AI draws from for rates, hours, locations, accepted items, customs procedures, and FAQs.
Automated Notifications
Email confirmations triggered automatically when packages are tracked or status changes. Customers and staff stay informed without manual effort.
MCP Integration
Model Context Protocol (MCP) server connects the AI agent to live tracking data and business tools via a clean, extensible interface.
How It Works
The system has four integrated components that work together seamlessly:
Customer Chatbot
Website chat widget powered by AI. Receives natural language queries and sends them to the intent classifier.
Intent Engine
AI classifies the query type — track, history, billing, or QA — and routes to the correct handler.
MCP Server
Exposes tracking and counting tools via Model Context Protocol. The AI agent calls these for live package data.
Knowledge Base
Curated business information — rates, hours, customs procedures, accepted items — served to the AI for general queries.