Hello World!

I’m Larry, a Forward Deployed AI scientist at Boston Consulting Group based in the NYC office. I’m experienced in personalisation (uplift modelling, response modelling, model deployment) and GenAI (rag QA chatbot, pretraining). I was part of the DeepAI team where I built a personalization solution that delivered over 10M in projected incremental revenue and >20% uplift in upsell rates. I also open sourced a tabular ML starter repo used by DeepAI which allows projects to deliver impact quickly and at scale. Link here.

Previously, I was a team lead at DSO National Laboratories working on pretraining large language models (70B models on >200 GPUs). I’m most proud of leading the pretraining team (with no experience) to pretrain a model that’s competitive with SEA AI Lab. There, I also contributed to the open-source pretraining library TinyLlama.

I graduated from National University of Singapore Computer Science with a first class honours. During my time there, I tried out research on NLP and bayesian optimization but found out I’m a better engineer than I am a researcher.