Ignacio Zañartu Höppner

I’m a Computer and Information Engineering student focused on Data Science, Artificial Intelligence, and Machine Learning. I build deep learning and reinforcement learning models in Python, with PyTorch at the core of my workflow.

In my featured project, I developed an Artificial Intelligence agent using Deep Q-Network (DQN) that autonomously learns to solve levels in a 2D environment (gym-super-mario-bros). I applied reinforcement learning fundamentals like experience replay, a target network, and an epsilon-greedy policy, then trained with GPU support and evaluated using metrics such as completion rate and convergence.

I care about engineering quality as much as model performance. I compared results against a baseline agent based on a Finite State Machine (FSM), and I implemented the project with modular architecture, Git-based version control, and automated testing with pytest.

I’m also intentional about reproducibility and learning from experiments—designing repeatable setups so results can be validated. Beyond projects, I’ve mentored new students and taught a practical video editing course, and I’m now looking to join remote or in-person teams building data-driven solutions.