TELL US ABOUT YOUR ROLE AND MAIN RESPONSIBILITIES.
Nick: As the Data Science Director for Natural Language Processing (NLP) and Generative AI (GenAI) at American Express (AMEX), I am responsible for leading the research and development in the application of cutting-edge NLP, GenAI and large language models (LLMs) to solve business challenges and enhance the customer experience.
Yingzhu: I am a Data Scientist specialising in NLP and GenAI. My responsibilities include leveraging the latest NLP techniques to reduce credit and fraud risk and deliver the best customer experience. My role entails end-to-end efforts — from understanding the business problem, to data collection and processing, and finally model development and deployment. I also work with cross-functional teams to understand their business needs and ensure that the model I deliver meets their needs.
WHAT MADE YOU PURSUE A CAREER IN THIS SECTOR?
Nick: As a global financial hub, Singapore offers abundant opportunities and job security in this sector. As a data scientist, I am drawn to industries with high-quality, diverse data. Financial institutions manage vast amounts of valuable data, presenting exciting opportunities to develop advanced AI solutions. The combination of job stability and access to rich data sources makes this field especially appealing.
Yingzhu: I find it rewarding to work on products that enhance consumer experience and promote financial literacy. I am also excited about staying ahead in a rapidly evolving market, where data help businesses formulate a data-driven response to manage risk in a fluid environment. On the technical side, GenAI and NLP technologies are at the forefront of innovation that enable me to tackle complex challenges.
WHAT DO YOU ENJOY MOST ABOUT THIS ROLE?
Nick: I enjoy AMEX’s collaborative culture, which offers a platform to learn from various business functions and build AI solutions for real-world needs. This collaboration enhances my understanding of the finance industry through expert insights and brings great satisfaction as I see the positive impact of my data-driven solutions — whether in improving operational efficiency or enhancing customer experience.
Yingzhu: I find it rewarding to transform raw data into insights that drive decisions. Explaining my complex model to stakeholders and seeing it perform well in volatile conditions brings great satisfaction. AMEX’s collaborative culture and the opportunity for continuous learning in this evolving field further fuel my enthusiasm at work.
WHAT WERE SOME WORK CHALLENGES YOU FACED AND HOW DID YOU OVERCOME THEM?
Nick: One of my main challenges is recognising that the best model is not necessarily the one with the highest accuracy, but the one that users adopt and delivers tangible value. It is essential to look beyond numbers and understand the business needs. I prioritise frequent communication with stakeholders during the model-building process to ensure alignment on solving the right problem. This approach also helps ensure that the final solution integrates seamlessly into users’ workflows. Through this close collaboration, I deliver data-driven solutions that are both technically sound and practical for real-world applications.
Yingzhu: I encounter many challenges in my day-to-day job, such as ensuring data quality for analysis and model building. To address this, I first focus on understanding and validating data sources, followed by implementing robust data cleaning and preprocessing techniques. Staying updated on the latest AI developments is also essential; dedicating time to continuous learning and participating in journal clubs helps me remain current.
WHAT IS NEEDED TO BE SUCCESSFUL IN THIS ROLE?
Nick: A strong commitment to continuous learning is crucial in this rapidly evolving field. Staying curious, keeping up with advancements, and enjoying learning new tools and techniques are essential. Equally important is taking ownership of projects — understanding the problem deeply, developing practical solutions, and ensuring effective real- world implementation. Success in this role comes from seeing the bigger picture and delivering solutions that integrate seamlessly into business processes.
Yingzhu: You need to be curious, persistent and collaborative. You should be willing to devote time to understand the business problem, cleaning and preparing data. Additionally, effective collaboration across functional teams ensures successful delivery of your project. One key attribute to an effective collaboration is effective communication, especially in translating technical concepts to a general audience.
SHARE YOUR ADVICE WITH STUDENTS WHO ARE KEEN TO PURSUE A CAREER IN YOUR FIELD.
Nick: For students interested in pursuing a career in data science, it is important to know that real-world work is more than building models. A significant amount of time is dedicated to data cleaning, understanding business needs, designing solutions, and collaborating with tech teams for deployment. It is not just about the model — it is about solving business problems and delivering actionable solutions. I recommend doing relevant internships to gain practical experiences and a better understanding of the day-to-day responsibilities.
Yingzhu: You need to be proficient in various Machine Learning and Deep Learning algorithms, choosing the right tool for each problem. Developing strong communication and collaboration skills is crucial to ensure your model addresses the business problem and secures partner “buy-in.” Lastly, networking with professionals in your field provides invaluable insights and opens doors to new opportunities.
