Business Problem

Context

Modern hospitality environments — like smart hotels — rely heavily on a wide range of internet-connected services: guest streaming, secure transactions, IoT devices (smart locks, lighting), operational systems, and public Wi-Fi. With encrypted traffic becoming the norm (due to HTTPS and VPNs), traditional monitoring systems based on Deep Packet Inspection (DPI) are no longer sufficient or scalable.

Problem Statement

Network Operations teams lack real-time visibility into what types of traffic are flowing across their network, especially when traffic is encrypted.

This leads to:

• Inability to prioritize critical services (e.g., POS, security systems)

• Difficulty detecting bandwidth-hogging or malicious usage

• Poor Quality of Service (QoS) for high-value guests or core applications

Traditional tools either:

• Cannot classify encrypted traffic accurately, or

• Require expensive appliances and invasive inspection

ML-Based Solution

This project introduces a machine learning model that classifies network traffic into meaningful application categories