Mar 2026 – Present · 2 mos

• Formally appointed by NUHS Deputy Chief Executive and Group Chief Digital Officer to lead technical delivery of a cluster-priority agentic AI system.

Apr 2025 – Present · 1 yr 1 mo

• Technical development lead for the 10-person AI SWAT team within the Digital Think Tank, focused on rapid AI prototyping and production deployment. Scope extends to cluster-level agentic LLM development through the NUHS AI Prioritization Workgroup.
• Technical Architecture
• ∙ Full-stack ownership across NTFGH’s AI ecosystem: EMR integration (Endeavour AI), production RESTful APIs, Kubernetes-orchestrated deployments on HCC AWS, with multicloud experience
• ∙ Introduced production-grade engineering standards: version control, automated linting, unit testing, OpenTelemetry observability, and LLM monitoring (LangSmith/MLflow)
• ∙ CKAD certified; AWS Certified Generative AI Developer - Professional.
• Clinical AI Systems
• ∙ Built cluster-wide ED audit automation saving 100+ manhours annually per ED (G.R.O.S.S. Commendation Award); now developing agentic pre-rounding summarization targeting 60-70% reduction in preparation time
• ∙ Led AI Avatar from POC to active development: STT-LLM-TTS voice pipeline, Three.js rendering, React/TypeScript frontend for patient admissions
• ∙ Designed LLM evaluation frameworks covering prompt optimization, clinical accuracy validation, and output consistency testing for production reliability
• Strategy & Leadership
• ∙ Guided team through 5 AI prototypes in 2025, with first now in production
• ∙ Technical consultant to senior leadership on AI architecture and build-vs-buy decisions
• ∙ Delivered executive presentations on AI strategy to C-suite and department leadership
• ∙ NTFGH representative at NUHS LLM PMO, contributing to cluster-wide AI collaboration strategy

Sep 2023 – Apr 2025 · 1 yr 7 mos

• Individual contributor role building foundational clinical AI capabilities—infrastructure, prediction models, and NLP pipelines.
• Infrastructure & Engineering
• ∙ Built production clinical web apps on HCC AWS using Streamlit, Docker, and Kubernetes
• ∙ Automated AWS resource provisioning with Terraform; integrated ECS/EKS for containerized deployment
• ∙ Developed SQL extraction scripts for Epic Clarity database to support research and predictive modeling
• Clinical AI & NLP
• ∙ Developed clinical prediction models (ICH Deterioration Risk, Length of Stay) using PyCaret AutoML and SHAP for explainability
• ∙ Built prediction logging and performance monitoring for production model reliability
• ∙ Engineered clinical NLP pipelines using LLMs, BERT, and MedSpaCy for NER and text mining
• Research & Impact
• ∙ Lead developer on project securing $100k+ in research grant funding
• ∙ Conducted statistical analyses (propensity score matching) for clinical program evaluations.
• ∙ Co-authored peer-reviewed research on clinical prediction models (intracerebral hemorrhage, emergency laparotomy pathways)
• ∙ Attained 5 Epic certifications and 2 AWS Practitioner certifications
• ∙ Led AI/ML technical workshops and mentored junior analysts