Pai Liu
Head Of Data Science, TikTok Shop Americas & Global Product And User Growth
San Francisco Bay Area United States
Summary
🤩
Rockstar🎓
Top SchoolPai Liu is a data science leader with 10 years of experience building and scaling ML-driven product and growth organizations across consumer marketplaces. As Head of Data Science for TikTok Shop Americas and Global Product & User Growth, she led a 150+ team and helped grow US TikTok Shop to $100M GMV/day in 18 months while developing LLM-based productivity tools. Previously she built Wish’s data organization and experimentation platform as CDO and led search and personalization teams at Airbnb, shipping ranking and recommendation models that expanded Airbnb into a broader travel marketplace. Trained as a PhD-level researcher in industrial and systems engineering with a master’s in computer science, she blends rigorous experimentation, ML engineering (notably contributions to Airbnb’s aerosolve repo), and product strategy to turn complex data into measurable business impact.
10 years of coding experience
9 years of employment as a software developer
PhD Industrial and System Engineering, PhD Industrial and System Engineering at University of Southern CaliforniaBachelor's degree, Bachelor's degree at Tsinghua University
English, Chinese
Github Skills (8)
machine-learning10
spark10
evaluation10
metric10
scala10
python9
linear-models9
decision-tree8
Programming languages (1)
Github contributions (2)
airbnb/aerosolve
Feb 2016 - Aug 2016
A machine learning package built for humans.
Role in this project:
ML Engineer
ML EngineerContributions:18 commits, 11 PRs, 12 pushes in 5 months
Contributions summary:Pai primarily contributed to the machine learning aspects of the `aerosolve` repository, focusing on the `FullRankLinearTrainer` and evaluation metrics. Their contributions involved implementing and refining different loss functions, specifically the squared hinge loss. They also added and refined evaluation metrics like Mean Average Precision (MAP) and Precision/Recall AUC, enhancing the model's performance assessment capabilities. Furthermore, the user introduced configuration options like `max_features` to customize the decision tree and forest trainers.
for-humanspythonmachine-learningdata-science
Patrick1860/code_snippet
Mar 2018 - Mar 2018
Contributions:1 push, 1 branch in 1 day
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.