Chi-Hsuan Chang

Hello! I am Chi. I’m a hands-on data science leader, passionate about diving into projects and getting things done.

My quantitative strengths lie in statistics and machine learning, including experimentation methodologies, value metric designs, and the productionalization of machine learning models on real-time A/B testing platforms. I specialize in uplift modeling, Bayesian statistical modeling, and more. Recently, I’ve been expanding my expertise into natural language processing, deep learning, and large language models (LLMs).

I’m experienced in leading data science teams and digital product development, with notable achievements such as:

• Experimentation and ML personalization (e.g., omnichannel, content, timing optimization)

• Identifying high-value customer segments using propensity and uplift scores

• Developing Python-based campaign performance measurement frameworks using frequentist and Bayesian methods

• Customer profiling, including defining patient cohorts with ICD-10 code logics and machine learning

• Software development (both front-end and back-end) of forecasting models (rule-based and ML-based) on cloud platforms

• R&D of generative AI agents using LLMs and retriever-augmented generation (RAG)

During my graduate studies, I was an environmental health researcher and biostatistician at Harvard, focusing on environmental pesticide exposure and health risk simulation using statistical models. My doctoral dissertation explored the link between sub-lethal pesticide exposure, honeybee colony collapse disorder (CCD), and human health using novel biomarkers.