Projects
🔹 AI-Driven Biometric Intelligence: Gender Prediction via CNN-Based Fingerprint Analysis
This project explores the application of Convolutional Neural Networks (CNNs) for gender classification from fingerprint images.
By leveraging deep learning architectures, it aims to uncover patterns that distinguish male and female fingerprints, offering potential applications in forensic science and biometric security.
🔹 Cyber Threat Intelligence Telegram Bot - Credential Risk Assessment Tool
Developed an AI-driven tool to assess the risk associated with credentials detected by a Telegram scraping bot.
- Conducted in-depth data engineering research, including feature selection, data collection, and preprocessing to build a high-quality dataset.
- Designed, trained, and evaluated a Machine Learning model to enhance risk assessment accuracy.
- Deployed the AI tool in a web-based triage system, enabling automated risk classification and visualization of credential threats.
Presented the project at the CNN-CERT (National Cryptologic Center) Cyberdefense Conference
🔹 Malicious PDF Detection Using Autoencoders Trained on Metadata
This project explores the use of deep autoencoders for detecting malicious PDF files through anomaly detection based on metadata.
The goal is to build a lightweight and interpretable model capable of flagging suspicious documents without requiring labeled malicious data during training.
🔹 Fire Detection using IoT Sensor Data with KNN and Decision Tree
This project compares the performance of K-Nearest Neighbors (KNN) and Decision Tree algorithms in predicting fire incidents using IoT sensor data.
The objective is to determine which of these Machine Learning algorithms provides better results for classification tasks
🔹 Smart Home Security System using Raspberry Pi
System designed to enhance home security and comfort by integrating real-time monitoring and automated alerts.
Built on a Raspberry Pi it implements a multi-sensor system with real-time monitoring, automated alerts, and a web-based interface for remote access and data visualization