In this hands-on presentation, learn how Python, Kubernetes, and practical programming came together to build an AI capable of beating the iconic 1990s arcade game, Double Dragon. Instead of complex AI theory, this talk shares an accessible story about how developers can leverage familiar tools to achieve remarkable results.

You'll see how I programmed an AI to interact directly with the game using PyBoy, a Python-based emulator specifically designed for seamless integration with Python scripts. Highlights include:

• Creating and deploying the PyBoy emulator within Kubernetes, enabling scalable and repeatable AI training sessions. I'll demonstrate containerizing PyBoy, managing Kubernetes resources efficiently, and ensuring consistent, reproducible training environments.

• Designing effective reward systems in Python to guide the AI toward mastering complex game scenarios. We'll dive into how reward structures were crafted to incentivize strategic gameplay behaviors, ensuring the AI learns efficiently and effectively.

• Highlighting the AI’s victory, showcasing in real-time how it outperformed human players. You'll witness live demonstrations of the trained AI conquering increasingly challenging game scenarios, demonstrating the power and practicality of integrating Python and Kubernetes.

This session offers relatable insights into how developers can practically implement AI in projects using open-source tools and accessible programming practices.