Full-stack · Machine learning
Multimodal Phishing Detection
A full-stack web analysis platform that connects a React interface with Express and Flask services for URL and DOM analysis and machine learning classification.
The overview
Listed in the résumé as the Multimodal Web Intelligence & Analytics Platform, this project combines a responsive web application with automated extraction and machine learning classification services.
The problem
Web analysis spans multiple inputs and processing steps. The application brings those services into an interface where a user can submit a URL and review the returned analysis.
The objectives
- Connect a React frontend to Express and Flask services.
- Process URL and DOM payloads through extraction pipelines.
- Present classification outputs through an understandable web interface.
How it comes together
Frontend and service integration
Developed a responsive React frontend backed by an Express/Flask microservice architecture and REST APIs.
Analysis pipeline
Integrated heuristic extraction and machine learning classification endpoints. The project stack includes Python, ResNet-50, and LightGBM.
The outcome
What the work delivers
- An integrated web interface for submitting and reviewing analysis.
- Connected extraction and classification services with user and admin views.
A closer look
The interface in action.
Have something similar in mind?
Let’s build it together