Zoe Yeh Email & Phone Number

Taiwan

Directed the establishment of a data-driven culture at HackMD as the first data team member, collaborating with C-level executives and the product management team to develop data infrastructure, pipelines, and dashboards.- Designed and implemented a data transformation system from scratch integrating databases and user logs, enhancing data accessibility and streamlined analytical processes, while optimizing pipelines to achieve a 40% cost reduction- Conducted hypothesis testing for the product comment feature update - Led the development of product metrics from scratch in collaboration with product managers- Developed customer segmentation data modeling by Python scripts to improve email open rate- Created 2 OpenSearch Dashboards for monitoring daily product feature adoption rate- Optimized the data pipelines and achieved approximately 40% cost-down- Presented at the 2024 Women in Data Science (TWiDS) conference on leveraging data analytics to identify potential paying customers, covering user behavior analysis and product analysis.

- Developed 13 project proposals, including MLOps, NLP, OCR, CV, Time Series, and Recommendation Engine.- Conducted 7 assessments and cloud architectures across various sectors.- Successfully executed 1 POC involving teams from Singapore and Taiwan, working with Google Engineers to build data pipelines and ML platforms.

- Delivered impactful data analysis solutions across 10+ projects on AWS and Azure, empowering clients with self-service analytics capabilities.- Developed and implemented an end-to-end data analysis platform for the Hong Kong Financial Industry, resulting in a projected 32% boost in query performance.- Led the implementation of NLP applications & MLOps platform (Natural Language Processing & Machine Learning Operations) for the Hong Kong Financial Industry, achieving the following: * Built NLP models to analyze social media sentiment, aiding in financial loss prevention. * Established an MLOps platform for automated model retraining, streamlining model management.- Spearheaded the implementation of a text search engine to optimize self-service analytics and significantly improve text search efficiency within the Hong Kong Financial Industry- Developed a computer vision (CV) model and MLOps pipeline for the Hong Kong Public Sector. This model detects rail defects, contributing to the prevention of major public safety accidents (daily metro passenger flow exceeds 500,000).

研究所, 統計研究所

大學, Mathematics