Data Science and Machine Learning Engineering - Leading ML model development and deployment- Built resource optimization model for cloud gaming service system- Delivered QoE (Quality of Experience) predictive model of cloud gaming service- Established data pipelines for ML development environment between data sources and platforms- Causal inference to understand user behaviors, QoE, etc.- Enhance experimentation environment and architecture.- Partnering with teams on DS and ML initiatives.- Collaboration with academia and research institutes- Tools: Databricks, AWS, Snowflake, Date Ocean, etc.

Annapurna Labs- Led customer research on SDK/documentation of deep learning models: computer vision (diffusion model, vision transformer, CNN, R-CNN), natural language processing (BERT, GPT, T5), LSTM, etc.- Managed and updated customer-facing documentation on AWS AI/ML accelerators (Inferentia andTrainium), improving the timeliness of content relevant to the release features- Delivered the new installation mechanism, effectively reducing QA testing process

Cloud gaming & research- Quality of Experience research- A/B Testing Platform- User engagement research- Game recommendation- Churn prediction model & Retention Strategy- Data-driven cloud game streaming optimization- Machine learning workflow (Kubeflow & Kubernetes)

Project: Sensor-aware Information Seeking Behavior Study- Research on information seeking behavior in different environments and contexts.- Designed and conducted experimental games (escape room and treasure hunt)- Collected GPS data via phone app and Google Maps Timeline (kml file).- Collected search logs via browser plugin.- Collected social interaction data (phone calls and SMS) via phone app.- Collected searching behavior in physical space via Google Glass.- Collected physical activities from Fitbit, via API.- Collected eye-tracking data on PC screen.- Analyzed data, writing code in Python, R, MySQL, and PHP.Project: People Analytics- Research project that investigates various signals from people, e.g., social media data, implicit andexplicit actions, as well as body sensors.- Managed research resources and developers to run sub-projects.Project: ContextMiner- Product that collects social media data -Twitter, YouTube, Flickr, and Blogs - for users.- Developed and maintained the code in PHP, MySQL, and Python.Project: SOCRATES- A product that collects social media data, visualizing the collected data and analysis results.- Social media of interest includes Twitter, YouTube, New York Times Comments, and Reddit.- Designed and developed the software framework in Python, PHP, JavaScript, etc.

- Software/hardware integration and validation of cable set-top box- Product & service development, e.g. home media server, widgets, and recommendation for customers- Presented products at CES (Consumer Electronic Show) and in-house demonstrations.- Programming in C and C++.