Projects

Multi-agent framework for detecting phishing emails using large language models. Implements MA2P, a system leveraging AI agents to identify malicious email threats through collaborative analysis.

Implemented a Deep Q-Network agent using PyTorch to solve the CartPole-v1 environment. Demonstrates core reinforcement learning concepts including experience replay and target networks.

Applied Quality-Diversity optimization to OpenAI Gymnasium's LunarLander-v3 environment using the pyribs library. The system discovers diverse behavioral solutions across the solution space.

Temporal difference learning agent that navigates a minefield environment. Implements value-based RL methods to learn optimal pathfinding while avoiding hazards.

Game AI implementation using minimax algorithm with alpha-beta pruning for optimal play. Demonstrates adversarial search and game theory concepts.

Algorithm implementation solving the classic chess puzzle where a knight must visit every square exactly once. Uses backtracking with Warnsdorff's heuristic.