At PSA INTERNATIONAL PTE. LTD., I delivered production LLM-based agents for industrial safety and process automation, from experimentation through deployment and user adoption. Multi-step workflows reached up to 95% accuracy on internal classification benchmarks, and concurrent LLM batch processing improved latency and compute cost efficiency by 80–90%.
I also built classification and analytics solutions for a B2B supply chain risk platform, including a daily monitoring tool for the Singapore Food Agency that achieved at least 80% recall and F1 score.
Beyond project contribution, on a corporate data/AI level, I initiated and published the company's first Databricks MLOps framework, a reusable reference architecture and set of accelerators for internal AI teams. It covers model lifecycle management, governance, monitoring, drift detection, and retraining pipelines.
My internships include experiences as the data analyst for a car-hiring start-up, timely delivering data analysis and reports to enable data-driven decision making in a fast-paced environment. I also have experience working on NLP problems for a FinTech start-up.
Earlier, I worked in process and project engineering at EXXONMOBIL ASIA PACIFIC PTE. LTD., where I led development of a logistics project that delivered $300K in savings.
Singapore