CV

Doctoral Researcher — LLMs for Structured Data and Knowledge Construction

University of Manchester

• Designed encoder- and decoder-based systems for inferring entity types, attributes, hierarchies, and semantic relationships from heterogeneous tabular repositories.
• Improved F1 and Recall by 30–40% and Rand Index by 7–10% over prior approaches across real-world datasets.
• Built modular Python pipelines with PyTorch, Hugging Face Transformers, and vLLM for batched inference, caching, structured validation, ablations, and reproducible evaluation.
• Developing an agentic multi-database question-answering system combining schema linking, structured reasoning, Text-to-SQL, and tool calling with Qwen2.5-14B.

Visiting Research Associate

Singapore University of Technology and Design

• Co-developed a unified benchmark for data-stream clustering, published in Proceedings of the ACM on Management of Data (SIGMOD 2023).
• Implemented CluStream, DenStream, DStream, and DBStream within a shared C++ codebase and reproducible evaluation framework.

PhD

University of Manchester

Computer Science

• Expected completion October 2026
• Dean’s Doctoral Scholarship

MSc

University of Edinburgh

High Performance Computing with Data Science

• Graduated with Merit

BSc

Tianjin University of Finance and Economics

Information Management and Information Systems

• GPA: 3.81/4.00; Top 5%
• Minor in Accounting

Schema Inference for Tabular Data Repositories Using Large Language Models

Taxonomy Inference for Tabular Data Using Large Language Models

Dataset Discovery and Exploration: A Survey

Data Stream Clustering: An In-depth Empirical Study