Researcher - Agentic Algorithms and Architectures

Huawei Technologies Switzerland AG, a leading technology firm that builds advanced computing architectures for artificial intelligence, is hiring three Researchers in Zurich. The position is a contractor role focused on extreme‑scale agentic computing, aiming to shape the next generation of AI systems through innovative research. In this role you will carry out both fundamental and applied investigations into agentic algorithms, modelling techniques, and hardware architectures. Daily tasks include designing new algorithmic solutions such as collective communications and resource‑scheduling strategies, constructing cost models and simulation tools to evaluate workloads, exploring modular interconnects and emerging technologies like near‑packaged optics or 3‑D stacking, performing systematic design‑space analyses, prototyping and validating concepts, and publishing results at top‑tier conferences. Close collaboration with hardware engineers, software developers and product teams is also required. The position demands a Master's degree or PhD in Computer Science, Computer Engineering, Electrical Engineering, or a closely related discipline, with a PhD being preferred. Candidates must demonstrate a strong interest in at least one of the listed research areas, algorithms, modelling, or architectures for agentic AI. These academic credentials and research enthusiasm are essential for contributing effectively to the team's objectives. Additional advantages include prior experience with AI or agentic workloads, which is considered a plus. The role is based in Zurich, Switzerland, offering a competitive salary and benefits package, as well as clear pathways for professional growth and development within a dynamic, cross‑functional environment. What the role asks for: - Master's degree or PhD in Computer Science, Computer Engineering, Electrical Engineering, or related field - Strong interest in at least one listed research area - PhD preferred - Experience with AI / agentic workloads (nice‑to‑have)