Kelvin has diverse research interests spanning several fields, including AI applications, data analytics, microlearning, financial education and financial technology. He has published peer-reviewed journal papers, encyclopaedia entries, book chapters and conference papers on different fields including Accounting, Banking, Business, Criminology, Computer Sciences, Data Analytics, Education, Engineering, Finance and Management. Currently, Kelvin's research focuses on exploring the applications of Large Language Models (LLMs) and computer vision. He investigates the use of LLMs in AI-powered recommendation systems and delves into the intricate relationship between LLMs and decision making. His work contributes to advancing understanding and utilization of LLMs in practical contexts. In computer vision, an application of his work that integrates AI with satellite data from the European Space Agency (ESA). This application has featured as a case study on the Copernicus Data Space Ecosystem, a platform under the Copernicus program. The Copernicus program is the European Union's most ambitious Earth observation initiative, which uses both satellite and ground-based data to tackle environmental and societal challenges on a global scale.

He has been involved in several funded research projects. For example, he served as the Principal Investigator for the project Evaluating Sentiment and Frequency Biases in GPT-4 Generated Financial Reports, which was funded by OpenAI’s Researcher Access Program and awarded USD $3,000 in API credits. In addition, he acted as a Co-Investigator on several other research initiatives, including Applying Fine-Tuned Artificial Intelligence (AI) Language Models to Enhance Stakeholder Engagement for Non-Profit Organisations, which received GBP £4,900 in QR funding from the University of Chester.