Delta Capture is an AI consultancy accelerating technology innovation in generative AI.

Delta Capture is an AI x Climate Tech startup helping financial institutions measure deforestation in commodity supply chains. Raised funding, assembled a world-class team of advisors, secured partnerships with KPMG & SAP, ran innovation and commercialization efforts of our AI technology to save Indonesia’s rainforest.

Preamble is a world-class startup spun out of MIT focusing on improving the safety of large learning models (LLMs).

SCV is the venture arm of Standard Chartered bank, driving industry-wide change through investments in fintech technology. Working with investors as a venture lead with ENGIE Impact, we derisked sustainability business opportunities in the intersection of sustainable finance, scope 3 decarbonization, and clean energy.

JAMM is an AI-powered insuretech company. Over the span of 16 months, we raised $800k+, assembled a world-class team of 9, and built a platform for safe driving with top insurance companies like DirectAsia, NTUC, and Etiqa—the safer you drive the more you’re rewarded. Product leveraged state-of-the-art AI/ML algorithms and was live on the road with over 100+ drivers in Singapore and the United Kingdom.

Entrepreneur First is a world leading talent investor. Investing time and money in talented and ambitious individuals, helping them to find a co-founder, develop an idea, and start a company. So far, EF has helped 2,000+ people create 300+ companies, worth a combined $2bn.

Artificial intelligence researcher: I was advised by Professor David Hsu at the National University of Singapore where my work focused on implementing a computer vision attention mechanism for modeling real-time sensory data for traffic prediction of autonomous vehicles. This was part of a larger research direction of towards an integration of planning and learning for robust AI decision-making in robotics.

I am advised by Professor Stefanie Tellex at Brown University, Department of Computer Science. My work focuses on the intersection of planning and collaborative robotics. Recently I have submitted a first-author paper to the International Conference on Robotics and Automation (ICRA) 2019 which involves solving a novel multi-object search (MOS) task. The robot uses an initial natural language command to set a prior over object locations (e.g. kitchen in the command "Please find the mugs in the kitchen"). The robot then solves the task with a general online planner, called OO-POMCP, which was developed in the paper. One important aspect of this work is supporting robust planning in partially observable environments. Future research hopes to further advance the social-cognitive role of robots in personal or industrial environments for mobile manipulation, assembly, or navigation domains.

My research included two projects: 1). Developed a scalable algorithm for retrieving relevant information from a semantic network via probabilistic reasoning over a cognitive agent’s current knowledge. 2). Leveraged the structure of a task to set cues to store, maintain, and retrieve informationautonomously so as to optimally reduce working memory of a cognitive agent.All lab operations are conducted within SOAR cognitive architecture, under advisor Professor John E. Laird (University of Michigan, Department of Computer Science).

Built and taught my own university registered honor’s course for Freshman, which examined how various disciplines study consciousness.

Supervised the operations of managing a team of editors, reviewing student articles, and publishing them. Alone raised $5,150 dollars for my organization in 3 months.