PhD in Artificial Intelligence Algorithms

Key Responsibilities

China Unicom (Hong Kong) Operations Ltd is a member of China Unicom that provides telecommunication services in Hong Kong. To cope with our development, we would like to invite high caliber candidates to apply for the following position:

• Research & Development:
• Design and implement advanced algorithms for Multi-Agent Reinforcement Learning (MARL), including:
• State representation methods for heterogeneous agents based on Graph Neural Networks (GNNs).
• Differentiable credit assignment mechanisms (e.g., improved algorithms based on Shapley values).
• Sparse communication protocols for partially observable environments (e.g., event-triggered message passing).
• Integrate World Models for environment dynamics prediction to reduce training costs.
• Optional: Explore human-AI collaboration through natural language explanation modules (e.g., decision summaries generated by large language models).
• Theoretical & Experimental Validation:
• Prove the convergence of proposed algorithms (e.g., theoretical analysis under submodular tasks).
• Evaluate algorithm performance in simulated and real-world scenarios (e.g., robotics, logistics).
• System Optimization:
  
  

Develop lightweight deployment solutions (e.g., model distillation, edge computing).

• Collaboration & Dissemination:
• Strive to publish research results in top-tier conferences and journals.
• Collaborate with engineers to translate research into practical prototypes.
• Other duties as assigned by the department manager.
  
  

Qualifications

• Required:
• A PhD (completed or near completion) in Computer Science, Artificial Intelligence, Robotics, or related fields. Priority will be given to fresh PhD graduates
• Strong foundation in Multi-Agent Systems, Reinforcement Learning, and Deep Learning.
• Hands-on experience with MARL frameworks (e.g., QMIX, MADDPG) and Graph Neural Networks (GNNs).
• Proficiency in Python, PyTorch/TensorFlow, and simulation tools
• A track record of publications in AI/ML fields.
• Preferred:
• Experience in edge computing or model compression techniques.
• Strong computational background with extensive production-level programming and scripting experience.
• Solid experience in machine learning or deep learning.
• 1+ years of software engineering and AI/ML experience is preferred.
• Practical experience with Python or C++.
• Experience building and/or delivering prototypes on API platforms.
• Familiarity with container technologies (e.g., Docker, Kubernetes), knowledge of MLOps and related technologies/methods.
• Experience supporting machine learning and AI applications with cloud technologies, understanding of DevOps tools.
• Hands-on experience with CUDA and CANN is a plus.