CV

Selected research, engineering experience, projects, publications, and skills.

Contact Information

Name Yufeng Huang
Email yufeng.huang.ai@gmail.com
Website https://www.dreamhungry.com/

Professional Summary

Game AI and tools engineering lead with 8+ years of industry experience at Tencent, ByteDance, and NetEase. My work spans behavior-based NPC intelligence, LLM-driven agents, AI-assisted production workflows, and generative AI for games. My research sits at the intersection of AI for games, intelligent interactive agents, and human-AI co-creation, with a focus on agents that augment human game designers throughout the creation process.

Experience

  • 2024 - Present

    China

    Tools Engineering Team Lead & Game AI Lead
    Tencent, SGRA Studio
    Lead the studio’s tools engineering team and serve as its Game AI lead, overseeing development tools, AI-assisted production workflows, and game AI initiatives.
    • Designed and built a hybrid NPC AI architecture for Animula Nook, combining behavior trees, finite-state machines, utility AI, and LLM-based modules. Dozens of fully autonomous NPCs operate concurrently per scene, while behavioral level of detail and resource scheduling support millions of NPC entities across a persistent shared world.
    • Led development of 10+ LLM-powered authoring tools for narrative and dialogue generation, behavior tree generation from natural language, level generation, and automated cutscene generation, substantially reducing content creation time.
    • Built an experimental pipeline in which an LLM agent autonomously generates production tools from UI design through code implementation and automated testing, while exploring multi-agent approaches to general-purpose game generation.
    • Contributed to core gameplay systems including character animation, NPC systems, dialogue, quests, combat, and in-game chat.
    • Designed and built the globally distributed server architecture for Animula Nook, supporting 1M+ DAU in a single shared world.
  • 2021 - 2024

    China

    Senior Software Engineer
    ByteDance, Nuverse
    Worked at Nuverse from June 2021 to December 2023 and from February to July 2024, developing game AI systems and production tools across multiple studio projects.
    • Built the battle AI for Dragonheir: Silent Gods, a launched strategy RPG with millions of downloads, using behavior trees and state machines for auto-chess-style tactical combat. Explored reinforcement learning for hero lineup composition and positioning strategies.
    • Designed complex multi-phase boss AI using hierarchical behavior trees and state machines, modeling player-adaptive attack patterns and phase transitions.
    • Built an event-driven, blueprint-based NPC AI system for Dragonheir: Silent Gods, supporting 1,000+ concurrent NPCs per scene and extensive scripted narrative performances.
    • Developed a utility-based ambient NPC behavior system for an internal open-world ARPG, enabling context-aware daily routines and environmental responses.
    • Explored generative AI in game production, including concept art and UI icon generation with diffusion models, text-to-motion generation, and video-based motion capture.
    • Developed internal blueprint, behavior tree, narrative, quest, and dialogue editors, gaining direct insight into designer workflows and human-AI co-creation needs.
  • 2020 - 2021

    China

    Software Engineer
    NetEase Games, Dahua Studio
    Worked on an internal open-world MMORPG project, developing gameplay systems and companion AI.
    • Developed core gameplay systems including character control, camera, and level systems.
    • Built companion AI combining behavior trees and utility AI, enabling NPC allies to fight alongside players with context-aware combat behavior.
  • 2017 - 2019

    China

    Software Engineer
    NetEase Games, Dahua Studio
    Worked on Onmyoji Arena, a launched MOBA title.
    • Shipped multiple gameplay systems including hero skills, in-game activity systems, and battle pass systems.
    • Gained working knowledge of utility-based MOBA combat AI and participated in exploratory research on reinforcement-learning-based PvP agents.

Research Interests

  • Human-AI co-creation of games.
  • Interactive and autonomous virtual agents.
  • Multi-agent systems for creative and interactive tasks.
  • Game-playing AI and generative AI for games.

Education

  • 2014 - 2017

    Wuhan, China

    M.Eng.
    Huazhong University of Science and Technology
    Pattern Recognition and Intelligent Systems
    • Machine Learning and Data Mining
    • Pattern Recognition
    • Computer Vision
    • Matrix Theory
    • Stochastic Processes
    • GPA: 3.75/4.0.
    • Thesis: CNN-based Visual Place Recognition in Changing Environments.
  • 2010 - 2014

    Wuhan, China

    B.Eng.
    Huazhong University of Science and Technology
    Measurement & Control Technology and Instrumentation
    • C Programming
    • Data Structures
    • Computer Networks
    • Calculus
    • Linear Algebra
    • Probability Theory
    • GPA: 3.73/4.0.

Selected Research Projects

Publications

Skills

Programming: Python, C/C++, C#
AI / ML: PyTorch, Hugging Face Transformers, Reinforcement Learning, LLM Agent Development (LangChain, RAG, Context Engineering), Multimodal Interaction (ASR, TTS, Emotion Recognition)
Game AI: Behavior Trees, Finite-state Machines, Utility AI
Game Engines and Tools: Unity, Unreal Engine, Custom Editors and Authoring Tools
Backend and Infrastructure: FastAPI, WebSocket, Docker, Linux, Distributed Server Architecture