Projects

Experimentation Accelerator

Experimentation Accelerator is part of Adobe Journey Optimizer. It is an AI-first application designed to surface actionable insights and high-impact optimization opportunities.

Semantic Kernel Bandits for Gen-AI Powered Experiments

GenAI can generate massive amounts of content as treatments (texts/images/video/audio). We implement a kernel bandit algorithm that leverages the semantic information underlying the Gen-AI powered treatments to help A/B testing on Adobe.com.

BSTAR: Long-Term Ranking Dynamics Evaluation

What works in the short run does not always work in the long run. We developed a pipeline called BSTAR to align iterative ranker deployment decisions with strategic long-term goals.

rdborrow: Incorporating External Controls in RCTs with Longitudinal Outcomes

Designing and analyzing RCTs poses many challenges such as how to ensure the validity and improve the power for hypothesis testing with a limited sample size or how to account for a crossover in treatment allocation. We introduce a new R package called rdborrow, which implements several external control borrowing methods under a causal inference framework to facilitate the design and analysis of clinical trials with longitudinal outcomes.

Last edited on January 30^th 2026 11:15AM (Time Zone: PST).
Powered by jemdoc + new design.