AI researcher and MComp student at the National University of Singapore
About
I am a Masters student at the National University of Singapore pursuing a Masters in Computer Science, after completing my bachelors in Computer Engineering. My interests sit between systems, machine learning, reasoning, planning, and post-training of language models.
I am currently working at the Scalable AI Lab under Prof. Lu Yao, focusing on post-training LLMs into specialized knowledge-aware agents and building concrete benchmarks for measuring reasoning behavior.
I am open to AI engineering roles beginning in early 2027, specifically in LLM post-training, memory systems, and evaluation and benchmarking.
Projects
Planner-centric RL for deep research using typed DAG plans and structure-aware reward. DecomposeR-8B improves over open baselines by 5.1–8.0 points on long-form benchmarks.
U-Mem studies memory agents that actively acquire, validate, and curate knowledge through cost-aware escalation and semantic-aware exploration, improving QA and reasoning benchmarks.
Teaching small vision-language models to highlight what matters on your screen. 2B and 4B models, SFT + GRPO, a Chrome extension and 300 interactive test comparisons.
A 450-puzzle Rush Hour benchmark across 3x3, 4x4, and 5x5 grids, with puzzle images, text prompts, solution validation, and model performance tables.
A Singapore OSM visual route-choice benchmark where VLMs inspect colored candidate driving routes and identify the shortest route without using a routing API at inference time.
Experience
Infiniwell, Trondheim, Norway
Curium, Singapore
Technology Centre for Offshore and Marine Singapore
Education
National University of Singapore
National University of Singapore
Raffles Institution
Contact