Hi, I am Boyuan, an SE / AI Researcher.

👤 About Me

I am a Principal Researcher at Centre for Software Excellence, Huawei Canada.

My current research interests are building Software Engineering-first LLMs. Our team’s target is to build the most cost-effective SE models (Sonnet is great, but expensive).

In the past, I lead a research team working on Heterogeneous Computing. In particualr, we focused on building infrastructure for large scale clusters for FMWare (i.e., Software built on Foundational Models). I am also interested in software reproducibility, logging, and AI4SE. I received my Ph.D degree in Software Engineering from York University, Canada, under supervision of Prof. Zhen Ming (Jack) Jiang.

Many of our team’s work is contributed to Ray, a popular open source distributed computing framework. For instance, We provided the native Ascend support for Ray. Our talk at Ray Summit 2024 is here.

🆕 News

2025-03-28: We released our recent work on performance enhancement for AI-Native Coding. This work focuses on the Service Level Agreement of LLM serving for coding tasks. We present Coding Assistant Task Orchestrator (CATO), the first SLA-aware algorithms for LLM serving that’s been internally deployed into Huawei Cloud and on Ascend NPU clusters. CATO intelligently orchestrates CodeLLMs to meet diverse coding tasks’ unique latency requirements while maximizing resource utilization. Compared to model-centric serving approaches (e.g., Ray Serve), we have achieved up to 41.4% improvement on goodput. Check it out at here.

📚 Selected Publications (Full List)

• SLA-Awareness for AI-assisted coding.
  Kishanthan Thangarajah, Arthur Leung, Boyuan Chen, Ahmed E. Hassan.
  arXiv preprint arXiv:2503.19876

• Rethinking Software Engineering in the Era of Foundation Models: A Curated Catalogue of Challenges in the Development of Trustworthy FMware.
  Ahmed E. Hassan, Dayi Lin, Gopi Krishnan Rajbahadur, Keheliya Gallaba, Filipe Roseiro Cogo, Boyuan Chen, Haoxiang Zhang, Kishanthan Thangarajah, Gustavo Oliva, Jiahuei (Justina) Lin, Wali Mohammad Abdullah, and Zhen Ming (Jack) Jiang.
  International Conference on the Foundations of Software Engineering (FSE 2024).
• Rethinking Software Engineering in the Foundation Model Era: From Task-Driven AI Copilots to Goal-Driven AI Pair Programmers.
  Ahmed E Hassan, Gustavo A Oliva, Dayi Lin, Boyuan Chen, Zhen Ming (Jack) Jiang
  arXiv preprint arXiv:2404.10225
• Towards training reproducible deep learning models.
  Boyuan Chen, Mingzhi Wen, Yong Shi, Dayi Lin, Gopi Krishnan Rajbahadur, Zhen Ming (Jack) Jiang
  International Conference on Software Engineering (ICSE 2022)