Hi, I am Phong Do, a first-year PhD student in Computer Science at the University of Warwick, supervised by Dr. Gabriele Pergola.

My research focuses on building more reliable, robust, and adaptive LLM-based systems. I am especially interested in how large language models reason, retrieve evidence, follow instructions, and execute workflows in complex settings such as retrieval-augmented generation and agent-based systems.

Recently, my work has been moving toward LLM workflow engineering: how to represent, analyze, optimize, and adapt the reasoning structure of LLM systems. Instead of treating an LLM system as only one prompt and one answer, I am interested in the full workflow: planning, retrieval, tool use, model calls, verification, aggregation, and final response generation.

Before starting my PhD, I worked as an AI Engineer at Zalo AI, where I contributed to language and speech technologies for real-world Vietnamese AI applications. I also worked with the UIT NLP Group, where my research focused on Vietnamese NLP, machine reading comprehension, question answering, and language models.

My research interests

• LLM workflows and agentic systems
• Prompt robustness and prompt stability
• Retrieval-augmented generation
• Indirect prompt injection and LLM security
• Question answering and machine reading comprehension
• Multilingual NLP, especially Vietnamese NLP
• Knowledge graphs and structured reasoning

I am happy to chat and discuss potential collaborations. Please feel free to reach out to me via Email (phong.do@warwick.ac.uk).

📝 Latest Blogs

News