Data & AI Engineer | 10+ years shipping ML systems in financial services | Singapore
I build production systems that solve real problems. My work spans the full spectrum—from classical ML pipelines processing millions of transactions to LLM-powered agents that automate complex workflows. I care about systems that actually work in production, not just notebooks that demo well.
After a decade in banking and insurance, I've learned that the best solutions often combine battle-tested statistical methods with modern AI capabilities. The magic isn't in choosing one over the other—it's knowing when to use which.
Uplyft Tech — Co-founding an AI-native e-commerce platform
Building the intelligence layer for online retail. Our stack includes Synapse (multi-source data ingestion with AI enrichment), Beacon (semantic search and conversational shopping), and deep Shopify integrations. We're solving the problem of making e-commerce data actually useful—turning raw product catalogs and customer behavior into actionable intelligence.
Claude Code Ecosystem — Developer tooling for AI agents
Creating frameworks, plugins, and memory systems that extend Claude's capabilities. This includes persistent context management for long-running agents, specialized task agents, and development workflows that make building with LLMs more practical.
ML Infrastructure — Production pipelines at scale
End-to-end systems for model training, deployment, and monitoring. Focus on making ML reproducible, observable, and maintainable—the unglamorous work that separates POCs from production.
Languages: Python, R, SQL, Spark
ML/AI: Scikit-learn, XGBoost, LightGBM, PyTorch, LangChain, Claude API
Cloud & MLOps: AWS, SageMaker, Docker, MLflow, Airflow
Tools: Git, Jupyter, VS Code, dbt
