Computer Science @ McGill University | Game Development · Game AI · Reinforcement Learning
Last Updated: August 2026
Hi! I’m Eric Xu Cui, an undergraduate student studying Computer Science with a minor in Mathematics at McGill University.
I’m interested in game development, game AI, and reinforcement learning. Most of my recent work has focused on Unreal Engine, gameplay programming, and experimenting with learning-based agents in interactive environments.
I have worked on Unreal Engine projects through personal projects, teaching, academic showcases, and an industry internship.
Education: McGill University — B.Sc. Computer Science, Minor in Mathematics
Interests: Game AI · Reinforcement Learning · Unreal Engine · Gameplay Programming
Languages: C/C++ · Python · Java
Tools & Technologies: Unreal Engine 4/5 · Blueprint · Behavior Trees · Multiplayer / RPC · Reinforcement Learning · Json
Email: ericxucui.work@gmail.com
LinkedIn: Eric Xu Cui
GitHub: EricXuCui
Shanghai, China· Onsite · Summer 2026
During my internship at TANER GAMES, I worked on gameplay and AI-related features in Unreal Engine 5.
Some of the areas I contributed to The Awakener: Battle Tendency include:
NPC Dialog Interaction
Squad Companion AI Tree Implementation
McGill Undergraduate Science Showcase
This project explores the use of reinforcement learning for video game combat AI.
I built an Unreal Engine 5 environment where an agent observes the state of a combat encounter, selects actions, and receives rewards based on its interactions with an opponent.
The project was presented at the McGill Undergraduate Science Showcase.
Technologies: Unreal Engine 5 · C++ · Learning Agents · PPO · Reinforcement Learning
Photo taken at the McGill Science Showcase, March 2026
I am currently extending the Combat Learning Agent project to explore how real-time reinforcement learning agents decide when to make their next policy decision.
Different gameplay actions have different durations. For example, a dodge, light attack, or heavy attack may require different amounts of time before another decision is meaningful.
I am experimenting with action-dependent decision timing and interruption-aware scheduling to study how these mechanisms affect agent behavior and training.
This work is currently in development.
I developed a practical Unreal Engine course focused on implementing Unreal Engine networking and Remote Procedure Calls using C++.
The project covers:
I developed a practical Unreal Engine course focused on implementing combat-oriented game AI using C++ and Behavior Trees.
The project includes examples of:
The course is available in both Chinese and English versions.
This project explored motion-based interaction in Unreal Engine using C++, motion sensors, and multiplayer networking.
Player actions such as attacking and blocking can be controlled through physical device movement and orientation.
The project advanced to the semifinals of the BC Youth Innovation Showcase.
Presented the Combat Learning Agent project exploring reinforcement learning in a real-time combat environment.
Semifinalist for the Motion-Controlled Medieval Combat System project.
Developed Unreal Engine courses covering game AI, C++, Blueprint, and multiplayer programming.
I’m happy to connect with people interested in game development, Unreal Engine, game AI, reinforcement learning, or related projects.
Email: ericxucui.work@gmail.com
LinkedIn: Eric Xu Cui
GitHub: github.com/EricXuCui
© 2026 Eric Xu Cui. All Rights Reserved.