Research Associate

National University of Singapore • Singapore

Jun 2024 - present

Brief introduction:

AI Research Engineer – multimodal & knowledge-enhanced systems.

Responsibilities:

• Built production-oriented Graph RAG pipelines (vector DB, chunking, embedding optimization, retrieval tuning), which improved retrieval accuracy and consistence, and enhanced structured reasoning for clinical knowledge bases. (Paper under review)
• Conducted research on generative AI models (diffusion models, GANs) for synthesizing medical and time-series data, with my proposed FHRDiff model achieving an overall 75.3% performance improvement over prior state-of-the-art methods. Paper published
• Developed an automated OCR pipeline for extracting blood pressure monitor readings, reducing manual transcription workload and significantly improving clinical data digitization efficiency.
• Utilized Weights & Biases for experiment tracking and model evaluation, improving reproducibility and reducing debugging/training iteration time.
• Curated and preprocessed multimodal datasets (text, image, signal, and knowledge-graph data), enabling cleaner data pipelines and reducing preprocessing effort for downstream ML/LLM workflows.