Dr. Chen Lequn

Data Science Manager at American Express · Singapore

I am a Data Science Manager at American Express, where I focus on developing novel Generative AI applications in the financial services domain.

Previously, I was an AI/ML Computational Scientist at Accenture Singapore, where I architected and deployed production-grade Agentic AI platforms and multi-agent systems for enterprise clients. My work focused on building LangGraph-driven orchestration frameworks with agent-to-agent (A2A) communication protocols, enabling coordination across specialized agents for complex workflow automation in the financial services industry.

Before Accenture, I worked as a Research Scientist at the Advanced Remanufacturing and Technology Centre (ARTC), A*STAR, where I led projects on AI-driven manufacturing intelligence, including engineering drawing information extraction, CAD feature recognition, digital twin development, and real-time process monitoring.

I hold a Ph.D. in Mechanical Engineering from Nanyang Technological University (NTU), where my research focused on AI-assisted in-situ process monitoring and multi-sensor fusion for advanced manufacturing systems. I also received my B.Eng. with Honours (Highest Distinction) from NTU in 2021.

View CV Download PDF
Portrait of Dr. Chen Lequn

Expertise

My core strengths lie in bridging AI research and practical system deployment across enterprise environments.

Agentic AI & Multi-Agent Systems

  • Architecting hierarchical multi-agent orchestration frameworks with supervisor-worker mechanisms for complex workflow automation
  • Building production-grade agent-to-agent (A2A) communication protocols with asynchronous messaging and thread-based workflow tracking
  • Developing LangGraph-driven platforms with interrupt/resume capabilities for enterprise-scale coordination
  • Implementing multi-stage LLM pipelines with advanced prompt engineering for structured data extraction

Machine Learning & AI Infrastructure

  • Fine-tuning LLMs and VLMs (LoRA/QLoRA) for domain-specific applications
  • Building agentic RAG pipelines and GraphRAG systems with domain-specific knowledge graphs
  • Developing multi-modal ML systems combining vision, audio, and structured data
  • Deploying scalable micro-service architectures (Kubernetes, Docker, FastAPI) for AI systems

Manufacturing Intelligence

  • Digital twin frameworks for real-time process monitoring
  • CAD/CAM-based feature recognition and visual defect detection
  • Acoustic signal analysis for quality monitoring and anomaly detection

Visitors

World map with a flag for each country this site's visitors come from

Thanks for stopping by — each flag marks a country this site's visitors come from.