Projects
2025
PhishGuard AI: Intelligent Phishing Email Classifier
PhishGuard AI is a project I built from the ground up to tackle one of the most common cyber threats, phishing emails. Using machine learning and NLP, I trained models to accurately detect and classify phishing attempts in real time. The solution plugged into email servers, flagged malicious content and even educated users through a custom dashboard. With over 95% detection accuracy, PhishGuard didn’t just reduce alert fatigue, it gave SOC teams breathing room and visibility into evolving phishing tactics. This project sharpened my AI security skills and proved that smart automation is the future of threat defence.
2023
A Game-Theoretic Approach to Robustness of Routing
In this project, we explore how game theory can be applied to routing systems to improve their efficiency. We analyze the current state of routing algorithms and identify potential vulnerabilities, such as compromised nodes, that can lead to data loss and inefficiency. To address these issues, we propose a new approach that uses game theory to find the shortest and most secure path for data transmission. This approach involves representing each system in the network as a vertex, which makes decisions about how to forward the data. We also consider the impact of compromised nodes on the network and demonstrate how game theory can be used to find alternative routes that avoid these nodes. To evaluate our approach, we implement a simulation in Java that uses a graph representation of the network and determines the optimal path and Nash equilibrium. Through this study, we aim to demonstrate the potential benefits of using game theory in routing systems and provide a basis for further research.
2021
A Security Approach to Detect and Prevent ARP Spoofing using Active and Passive Techniques
Every device on a network talks using IP addresses, but the real action happens at the hardware level via MAC addresses. That’s where ARP (Address Resolution Protocol) comes in, matching IPs to MACs so data knows exactly where to go. But attackers can exploit this with ARP spoofing, tricking the network into sending data to the wrong place. To fight back, we use a mix of active and passive scanning to detect suspicious activity, and defensive techniques like frequently changing MAC addresses to keep intruders out for good.