Yao Gu Email & Phone Number

San Francisco Bay Area

Generative search quality

Greater Seattle Area

• Drove and implemented automated feature discovery services that focuses on launching personalized suggestions for Alexa customers using automation, which drove at least 4 millions incremental dialogs.
• Focused on intelligent personalized feature ranking system that smartly deliver Alexa recommendations to improve customer engagement.
• Designed and drove a contextual aware ranking feature, which support contextual signals for evolving the machine learning model to recommend content injection experiences.

Greater Los Angeles Area

• Developed robust responsive components on experience pages, including attraction products, reviews and tours, to improveuser experience and increase the conversion rate, involving Java with RESTful API, React, GraphQL
• Implemented, A/B tested and rolled out new features to improve experience, increase CVR and revenue.
• Optimized and engineered high performance endpoint services for low-latency query from multiple heterogeneous massivedata source including TripAdvisor and subsidiaries.

Marina Del Rey

• Human Trafficking Detection from Backpage
• Researched and developed robust information extraction, link prediction and entity resolution data pipeline for mining human trafficking from BackPage data using machine learning, graph embedding, to assist.
• Implemented a minimally supervised pipeline for mining millions real-time Twitter, by including a judicious mix of fast text semantic embedding, active learning.
• Paper: A Pipeline for Post-Crisis Twitter Data Acquisition (https://arxiv.org/abs/1801.05881)
• Improved the convergence performance and robustness of the pipeline and classifier by using a judicious combination of preprocessing and active learning. The recall was significantly improved by 20% than ordinary approach.
• Implemented an interactive Twitter semantic visualization tool, providing a rough estimate of semantic relatedness of massive Twitter dataset.

Suzhou, Jiangsu, China

• Bing Relevance & News Team, Search Technology Center Asia
• Implemented a text classification pipeline at Bing Index Engine for real-time text categorizing and Bing News question answering by using machine learning and NLP technology.
• Implemented news answering processor for high freshness required ranking. Implemented ranker to improve news answer relevance and quality.
• Designed and implemented high-performance MapReduce jobs to mine massive Bing Index dataset (hundreds GB daily) for model training and query answer analyzing.

Hong Kong

• Research Assistant Intern at Prof. James Cheng's laboratory.
• Implemented a robust distributed crawler by Python, using Scrapy, pyspider libraries and RPC.
• Crawled millions user movie review data for research on machine learning and super graph computing.
• Figured out various anti-crawler mechanism in practice by simulating real human behaviors.
• Optimized the network request by using dynamic proxy pool and optimized traffic flow using token-bucket algorithm.

Master, Applied Data Science

Bachelor Of Engineering - Be, Computer Science