I developed a Tic-Tac-Toe Neural Network AI, a Django REST API application providing an intelligent opponent powered by trained neural networks. Built with Django and Django REST Framework, this backend service offers a simple POST endpoint that returns strategic AI moves based on the current board state. The AI implements a multi-layered strategy combining rule-based logic with learned patterns, including win detection, block detection, center prioritization, and neural network-based move selection. The system uses two separate trained models for X and O players, each featuring a two-hidden-layer architecture with ReLU activation and softmax selection for optimal moves. Through reinforcement learning training, the AI has learned strategic gameplay patterns while maintaining robust input validation and error handling.