Zefei Xie

I am an M.S. student in Intelligent Information Systems at Carnegie Mellon University’s School of Computer Science. I completed my B.Sc. (Hons) in Computer Science with Artificial Intelligence at the University of Nottingham with First-Class Honours.

My work sits at the intersection of agentic AI, large language model systems, retrieval-augmented generation, human-AI collaboration, and data visualization. I am interested in building AI systems that expose their reasoning process, incorporate human direction, and produce reliable evidence-backed results.

Before joining CMU, I worked on multimodal model deployment and efficient inference at Zhejiang Dahua Technology, built internal applications at NetEase Hangzhou Research Institute, and served as a teaching assistant at the University of Nottingham.

Research Interests

Agentic AI and LLM systems: transparent, controllable workflows with tool use, retrieval, and iterative reasoning.
Human-AI collaboration: interfaces that help people inspect, steer, and trust AI-supported research.
Data analysis and visualization: systems that turn complex data and model outputs into actionable evidence.

Selected Projects

AwesomeLit: a human-AI system for literature exploration and research hypothesis generation.
IEEE VIS 2025 VISxGenAI Challenge: an agentic pipeline for autonomous data analysis and visualization that ranked Top 5 in the challenge.
IBM Watsonx AI Assistant: a retrieval-augmented assistant for new university students.

News

Jun 30, 2026	Completed my B.Sc. (Hons) in Computer Science with Artificial Intelligence at the University of Nottingham with First-Class Honours.
Mar 31, 2026	Our paper AwesomeLit: Towards Hypothesis Generation with Agent-Supported Literature Research was published as a EuroVis 2026 Short Paper.
Sep 01, 2025	Our paper was accepted in the IEEE VIS 2025 VISxGenAI Challenge and awarded with an oral presentation.

Selected Publications

AwesomeLit: Towards Hypothesis Generation with Agent-Supported Literature Research

Zefei Xie, Yuhan Guo, and Kai Xu

In EuroVis 2026 Short Papers, 2026

Abs DOI HTML

There are different goals for literature research, from understanding an unfamiliar topic to generate hypothesis for the next research project. The nature of literature research also varies according to user’s familiarity level of the topic. For inexperienced researchers, identifying gaps in the existing literature and generating feasible hypothesis are crucial but challenging. While general “deep research” tools can be used, they are not designed for such use case, thus often not effective. In addition, the “black box" nature and hallucination of Large Language Models (LLMs) often lead to distrust. In this paper, we introduce a human-agent collaborative visualization system AwesomeLit to address this need. It has several novel features: a transparent user-steerable agentic workflow; a dynamically generated query exploring tree, visualizing the exploration path and provenance; and a semantic similarity view, depicting the relationships between papers. It enables users to transition from general intentions to detailed research topics. Finally, a qualitative study involving several early researchers showed that AwesomeLit is effective in helping users explore unfamiliar topics, identify promising research directions, and improve confidence in research results.
What Happened to Automated Visualization? An Agentic Analysis

Weiyao Meng, Yuhan Guo, Zefei Xie, and 1 more author

In IEEE VIS 2025 GenAI Workshop, 2025

Abs HTML Code Website

An LLM-powered agentic workflow for planning, transforming, visualizing, and interpreting data with human-specified intentions and domain knowledge.