About the job
Huawei Canada is seeking a dedicated Researcher to join our innovative team.
At the forefront of technology, the Distributed Data Storage and Management Lab specializes in pioneering research in distributed data systems. Our mission is to create next-generation cloud serverless products, integrating core infrastructure and advanced databases. We tackle a myriad of data challenges, focusing on cloud-native disaggregated databases, pay-by-query user models, and optimizing low-level data transfers through RDMA. Our teams are committed to developing cutting-edge cloud serverless data infrastructures and implementing state-of-the-art networking technologies for Huawei's global AI ecosystem.
Position Overview:
Design and implement advanced Agentic Reinforcement Learning (RL) and Multi-Agent Reinforcement Learning (MARL) algorithms tailored for cooperative, competitive, and mixed-agent environments, incorporating CTDE, decentralized learning, and hierarchical agent systems.
Develop scalable simulation and training platforms for extensive agent systems, facilitating self-play, population-based training, curriculum learning, and emergent behavior analysis.
Enhance multi-agent learning performance across distributed compute clusters, focusing on sample efficiency, credit assignment, agent coordination, communication learning, and training stability.
Investigate and prototype innovative methodologies for multi-agent intelligence, encompassing communication protocols, credit assignment, game-theoretic learning dynamics, meta-learning, and adaptive agent populations.
Transform groundbreaking research in agentic AI and MARL into production-grade systems for real-world and high-fidelity simulated environments.
Establish benchmarking frameworks and evaluation metrics aimed at agent coordination, robustness, scalability, and safety.
Collaborate with research, infrastructure, and product teams to implement scalable agentic learning systems in practical applications.
Drive technical leadership and innovation through publications, patents, open-source contributions, and presentations at conferences.
The total target annual compensation for this role ranges from $106,000 to $156,000, contingent on education, experience, and proven expertise.

