Unimelb Reviewer | Hannah Zhao

Designed an AI-assisted academic review system to automate paper evaluation using structured criteria and LLM-based analysis.
Implemented a modular evaluation pipeline where different components assess aspects such as grammar, structure, novelty, citations, and relevance.
Integrated Retrieval-Augmented Generation (RAG) to generate context-aware feedback grounded in academic standards and reference materials.
Developed a hybrid approach combining rule-based checks with LLM outputs to improve consistency and reduce hallucination.
Built a prototype interface to support end-to-end review workflows, from paper input to structured feedback generation.
Collaborated with University of Melbourne and IBM teams to explore practical applications of LLMs in academic publishing.