Ansel Lin
Senior Data Scientist @ASUS
Signup · Get unlimited contacts
WORK HISTORY
Senior Data Scientist @ASUS
TW
AI Agent for Generative BI DeploymentBackground: PMs regularly relied on DA support for basic data queries, slowing down decision-making and resource allocation.1. Improved PM efficiency by 11% and QA accuracy by 30% by deploying an AWS AI Agent built with AgentCore, enabling cross-table data queries and knowledge base retrieval through system prompt and tool integration.2. Scaled solution from one country to 10+ regional teams, supporting faster decision-making across global e-commerce sites.Personalized Product RecommendationsBackground: EDM product suggestions were previously based only on trending items, lacking personalization and resulting in low click-through rates.1. Increased recommendation CTR by 19% by deploying personalized models via GCP Search for Commerce and AWS Personalize.2. Conducted A/B tests to compare model effectiveness across platforms and user segments, optimizing for contextual relevance.Predictive Targeting for EDM OptimizationBackground: Marketing teams relied on intuition for EDM targeting, leading to inefficient outreach and low campaign ROI.1. Improved EDM conversion rate by 7% through a GA4-based high-propensity model predicting purchase intent.2. Pioneered the team’s first end-to-end MLOps deployment — designed and implemented automated pipelines, retraining schedules, and drift detection using Vertex AI.3. Integrated GitHub Actions for CI/CD, increasing model reliability and enabling production-level stability.4. Designed and executed A/B testing frameworks to validate real-world performance of different modeling strategies.
EDUCATION
National Taiwan University
Master's degree, Economics
ABOUT ANSEL LIN
Bent on driving self-growth and delivering business value through data-driven and scalable machine learning systems. As a product-focused data scientist, I specialize in translating business goals into machine learning solutions that scale. I thrive at the intersection of data, product, and impact.At ASUS, I support 40+ global e-commerce sites by partnering with PM, CRM, and marketing teams to design RFM and high-propensity models using GA4 behavioral signals. I lead the end-to-end delivery of ML-powered solutions — from modeling to deployment — including automated pipelines built with Vertex AI and GitHub CI/CD. My GenAI projects span across search-based recommendations, content generation, and BI assistants, deployed on GCP and AWS to support global marketing and product operations.I built an AWS AI Agent for Generative BI using AgentCore, improving PM efficiency by 11% and QA accuracy by 30%. The agent enables cross-table data queries and knowledge base retrieval through system prompts and tool integration, and has been scaled from one pilot market to over 10 regional teams.I was the first in the team to implement scalable MLOps workflows from scratch, including retraining, drift detection, and cloud-based monitoring — and helped shape our team’s best practices for ML deployment and observability.My work has led to measurable results — boosting EDM conversion by 7% through a high-propensity model, increasing product CTR by 19% via search-based recommendations, and reducing PM insight turnaround by 11% using a generative BI prototype.Beyond modeling and deployment, I actively design A/B tests and propose uplift modeling frameworks to validate real-world impact — ensuring what we build not only predicts, but performs and scales.
This profile is compiled from publicly available professional sources. Unifers is not affiliated with or endorsed by LinkedIn. Request removal of this profile.