Search engineering

Major contributions: Grew a search-focused ML team and transformed it into a company-wide team handling all ML projects.* Led a machine learning team of 10, collaborating with global product managers and stakeholders to compile and execute the company's ML roadmap, driving strategic initiatives across multiple countries.* Led improvements on product search, including grouping similar products, query spelling correction, auto-complete, and ranking improvement, resulting in increased #views and click-through rates.* Led the team on homepage recommendation and personalization, diving into user behaviors.* Pioneered the use of OpenAI's GPT API for a chat-based product recommendation system.

Major contributions: Created a new search experience in a core team, serving as a backend engineer, machine learning engineer, and DevOps.* Built product search and comparison across various e-commerce sites in 6 countries, including Elasticsearch index building, datafeed processing pipelines, and recall/ranking relevance optimization.* Developed advanced search recall and ranking algorithms, achieving competitiveness with top market players like iPrice, through techniques such as BERT for product categorization and machine-learned ranking formulas.* Performed DevOps duties, managing operations on AWS/Kubernetes along with infrastructure engineers, to ensure robust, scalable, and reliable search services.* Created an internal tool that accelerated the ranking development cycle and improved confidence in offline test results, by simulating search traffic with visual result inspector.

* Launched a new real-time bidding (RTB) algorithm and gained better cost-per-action.* Reduced ~50% latency in the RTB AI runtime Python module for better throughput.* Refactored the dynamic recommendation system, including reducing the memory footprint to 50%* Led 3 engineers and worked with scientists to develop dynamic recommendation system and user segmentation AI, to give relevant content to relevant users.* Led the AI backend engineering team of six, to productionize training/inference works from scientists with CI/CD, monitor/alert, scalability, etc.

Major areas of contribution: software development, tech leading and data analysis* Applied machine learning techniques on　- Spam detection on user generated content in SVM and GBDT　- Machine-learned ranking across content providers, so users get the best federated search results in the first page.　- Implemented production runtime pipelines for model serving.* Led engineers of 3~6 to accomplish development of various products over 4+ years.　- Promoted Agile/Scrum to the team, for better transparency and communication among managers and engineers. Ensured good user experience end-to-end as well.* Participated in blueprint of Yahoo Search architecture.* Developed user-facing products, with experience across front-end, back-end, and index design:　- Unified vertical search experience (News, Sports, …) with multi-tenant technology in Java, so developers can onboard any vertical from all countries within weeks.　- Location detection in mobile web search, to deliver better results regarding locality intent* Promoted data analysis methodologies and technologies on Hadoop to the team, so everyone is able to make data-driven decisions.* Analyzed user behavioral metrics and provided strategies to improve product performance.

Developed a web-based system for microarray computation.

Developed a Java application monitoring sensors of power plants.