Applied AI · LLM Systems · AI Agents
I build practical AI applications with retrieval, tool use, evaluation and observable agent workflows — with a focus on systems that can be tested, inspected and explained.
I am pursuing a B.S. in Data Science and Big Data Technology at The Chinese University of Hong Kong, Shenzhen, with coursework spanning data structures, database systems, machine learning, software engineering, stochastic simulation, probability and statistics.
My current focus is applied AI: LLM APIs, structured outputs, agent workflows, execution feedback, embedding retrieval and practical tool integrations.
Built a feedback-driven testing agent that reads specifications, generates tests and metamorphic relations, executes checks, and uses failures to guide follow-up testing under explicit model-call and command budgets.
Built a production-style email workflow with confidence- and risk-based routing, evidence checks, low-risk FAQ automation, human escalation, traceable review history and durable send queues.
Hybrid BM25 + TF-IDF retrieval with reciprocal-rank fusion, inspectable source citations and a reproducible evaluation suite for grounded research workflows.
FinTech Department intern · Outstanding Intern. Researched AI-assisted scenarios for policy retrieval, document preparation and workflow coordination, and designed a multi-agent concept for branch operations covering customer service, employee assistance, risk/compliance, task orchestration, tool use and human confirmation.
Successful Participant. Contributed to mathematical modeling, data analysis, visualization and an English technical paper under time constraints.