Xi Luan
Hi, I’m LUAN Xi. I’m a multidisciplinary technologist blending computer science, machine learning, and game design to create engaging, data-informed experiences.
About
Hi, I’m LUAN Xi. I’m a multidisciplinary technologist blending computer science, machine learning, and game design to create engaging, data-informed experiences. I’m currently pursuing MSc programs in Computer Science and Informatics across institutions in London and Sweden, and I enjoy turning complex ideas into playable, polished products.
Beyond research, I love collaborating across teams to ship features and narratives that resonate with players while upholding strong technical and ethical standards.
- 1 year's experience
- Joined December 2025
Industry experience
- Gaming
- Media & Entertainment
- Software & Internet
- Education
- Professional Services
- Computers & Electronics
Work experience· 6 roles
Conducted applied research on diffusion-based models for pathfinding in 2D game environments, reframing navigation as a generative task. Designed and implemented datasets for maze trajectory generation (500 image pairs; 4,000 video clips) and built training/inference pipelines in PyTorch using Stable Diffusion, LoRA, AnimateDiff, and conditional UNet architectures. Achieved promising results with AnimateDiff’s text-to-video pipeline, generating visually consistent, goal-directed pathfinding sequences.
Conducted applied research on diffusion-based models for pathfinding in 2D game environments, reframing navigation as a generative task. Designed and implemented datasets for maze trajectory generation, including 500 image pairs and 4,000 video clips of path progression. Developed custom training and inference pipelines in PyTorch, leveraging Stable Diffusion, LoRA, AnimateDiff, and conditional UNet architectures. Achieved promising results with AnimateDiff’s text-to-video pipeline, generating visually consistent and goal-directed pathfinding sequences.
Built 5 levels per week aligned with the game’s vision and pillars; designed difficulty curves, level layouts, and pacing; incorporated feedback from senior designers to deliver engaging player experiences. Analyzed 1,000+ player logs per level post-release using SQL/Python, creating visualizations of progression trends and pinpointing high drop-off stages to drive design refinements.
Education· 12
Hire Xi Luan today
Post a job and invite Xi, or contact them directly. No hidden commission.











