Po-Wen Hsu

Hi friends! Welcome to Powen’s portfolio. 😊

I'm a data lover with 5 years of expertise, adept at leveraging data to drive product strategies, streamline project management, cut production costs, and boost marketing effectiveness. Specializing in semiconductor IC design houses, electronics manufacturing, and e-commerce.

I'm delighted to be contacted at email  /  LinkedIn  /  Github  /  Medium.
Currently, I'm open to new opportunities. Feel free to review my resume for further details. Additionally, here is a recommendation letter from my esteemed mentor, Jill, at Watts.

Something About Me

🍰 Passionate about food and travel
💡 A people person who loves embracing new ideas
🧳 Speak fluent SQL, Python, R and have Tableau and PowerBI on the side – consider me your data's multilingual tour guide

Professional Experience

- Product Process Analyst, Supply Chain @ Johnson & Johnson - Kenvue (Top 1 consumer health company)
- Product Analyst Intern, Web and Mobile Apps @ Watts (Major Player in the US Water System)
- Product Manager, Analog AI Integrated Circuit @ Egis (The second-largest fingerprint provider in the world)
- Product Manager, 400G QSFP-DD transceiver @ AOI (Primary transceiver supplier for data centers)

Projects

I have a passion for data science and am dedicated to refining my analytical skills to contribute to more informed decision-making in business strategies. My projects encompass a range of areas, including sentiment analysis, data mining, natural language processing, and the development of machine learning API applications, etc.

The Perceptron Learning Algorithm (PLA) is a fundamental concept in neural network training, providing the groundwork for understanding how single-layer perceptrons learn from input data.

Navigated through the intricacies of K-means clustering, unraveling the algorithm from scratch with Python. From a foundational understanding of the algorithm's inner workings to practical implementation.

Executed end-to-end data mining for house price prediction. Employed PCA for dimension reduction, Lasso for feature selection, and fine-tuned XGBoost hyperparameters. A comprehensive approach for accurate and efficient modeling.

Conducted a simple linear regression analysis on Columbia Business School professors using the statistical tool Minitab. The results yielded compelling insights, empowering the administration to enhance teaching quality.

Automated LinkedIn Job Insight. Seamlessly logging in, extracting the latest job listings based on preferences. Employing web scraping to extract detailed job descriptions and responsibilities, followed by a natural language analysis to identify key job position keywords.

Crafted a personalized café search engine in Python, leveraging the Google Places API and PostgreSQL. Integrated Spark for advanced attribute querying and incorporated Flask for a dynamic user interface, allowing users to find their ideal spot based on preferences, with real-time customization.

Exploring market dynamics through sentiment analysis reveals the surprising influence of Elon Musk's tweets on Tesla stock prices. Regardless of content tone, evidence shows Elon's tweets significantly impact Tesla's stock, uncovering intriguing insights into market behavior.

Enhance the Price and Occupancy Rate of Holiday Inn through a comprehensive quantitative analysis. Employ Tableau as a visualization tool to deliver a compelling presentation, showcasing data-driven insights and strategies for optimization.

Developed a research design plan and earned top honors in the final project presentation . Executed 1,000 simulated A/B testing studies, leveraging the R programming language to simulate a real-world scenario. This demonstrated a robust statistical approach and contributed to the project's success.

Unleashing the Power of Data to Optimize Advertising Strategy. Delve into innovative solutions, strategic insights, and data-driven approaches to propel advertising effectiveness and drive results. Secured Top 10 and Honorable Mention in the Amazon Hackathon Competition.