Professional Experience
Research and engineering experience in trustworthy AI, NLP, information retrieval, and healthcare.
Research Fellow
Jun 2024–PresentCentre for Trusted Internet and Community, National University of Singapore
- Developed trustworthy reasoning methods for NLP, focusing on model robustness under incomplete, noisy, and adversarial information.
- Proposed omission-aware fact verification frameworks that extend misinformation detection beyond binary truthfulness by modeling missing evidence, implied intent, and causal impact.
- Introduced reasoning-level vulnerability analysis for LLMs, showing that injected decision criteria can corrupt model judgments without changing the original task objective.
- Designed interpretable retrieval and diversity-aware evidence modeling methods for comprehensive event understanding under incomplete and biased information.
Research Assistant
Aug 2016–May 2024SeSaMe, N-CRiPT, and CTIC, National University of Singapore
- Developed Readpeer, a human-in-the-loop collaborative reading and annotation platform, improving annotation transfer, expert feedback integration, and document-level knowledge enrichment.
- Designed medical text mining algorithms for free-text electronic health records, extracting drug names, adverse events, and drug–adverse event relations from hospital discharge summaries.
- Worked with more than 10 clinicians, pharmacists, and regulatory experts to validate system outputs and support Readpeer-based drug safety monitoring in collaboration with Singapore's Health Sciences Authority.
Machine Learning Engineer Intern
Jun 2014–Aug 2014T.Ware Pie Ltd, Singapore
- Developed machine learning algorithms for human activity recognition using 3D acceleration data, improving accuracy from 60% to 85%.
- Developed an iOS application for intelligent stress-relief wearables.