About Me

I'm Amanda, a senior majoring in data science with a minor in Asian American studies and certificates in design and integrated marketing communications at Northwestern University. Whether I'm creating dashboards in PowerBI, Salesforce, and Tableau or developing technical solutions in Python and SQL, I tackle each project with enthusiasm and a hunger to learn and grow. I approach problems by surveying to understand what components are there and what has been done before. I’m also energized by working in teams and deeply value mentorship.

Between literature reviews on generative AI's implication in marketing and tutoring students in integral calculus, I am cooking rice bowls, people-watching, and creating 5-star Spotify playlists.

Projects

Determining Correlation Between Bi-Monthly Embeddings with Python XGBoost and Optuna

During my data science internship at Resonate, I developed on model-checking hopeful monster imputation.

  • Project Objective and Technical Focus: My primary goal was to assess whether the monthly waves of survey data could be differentiated post-imputation. I immersed myself in the intricacies of coding and statistical modeling, enhancing my understanding of these critical areas in data science.
  • Advanced Modeling Techniques: I utilized sophisticated machine learning techniques, including XGBoost for predictive modeling and Optuna for hyperparameter optimization. Additionally, I applied UMAP (Uniform Manifold Approximation and Projection) for effective dimensionality reduction and visualization of complex survey data.
  • Documentation and Communication: I meticulously documented all phases of the project, creating a comprehensive resource that outlined the modeling procedures and adjustments made. This documentation proved invaluable for the data science team, aiding in the continuity and clarity of the project.
  • Findings and Presentation: I discovered that the individual data waves were clumped together, a finding I presented to the VP of Data Science and AI. This presentation led to discussions on potential improvements and solutions for the imputation model.

Designing Demand Dashboard by Promotion by Marketing Channels

As an E-Commerce Analytics Intern at adidas, I developed CTE in SQL for a PowerBI dashboard by Promotion by Channel.

  • Data Schema Implementation and Preprocessing: Implemented and preprocessed a comprehensive data schema in Databricks using SQL for five datasets, encompassing over 390 million data points. This work supported the Data Solutions team in addressing Ad Hoc tickets from the North American (NAM) sports business units and the NAM e-Commerce buying team, ensuring accurate and timely data responses.
  • Self-Service Dashboard Creation: Designed and developed a self-service PowerBI dashboard that integrates data from consumer behavior, channel performance, promotions, and sales. This dashboard facilitates weekly updates on key metrics such as Average Order Value (AOV), demand trends, and top products by promotion, benefiting six channel teams as well as the business and experience team.

Enhancing User Experience for Older Adults on E-Commerce App

In Spring 2024, as part of a Digital Product Development course, my design group and I undertook a comprehensive project aimed at addressing the question: “How might an e-Commerce app better serve older adults?”

  • Research and Insights Gathering: To overcome the limitations of our initial assumptions, we engaged directly with older adults to obtain authentic and relevant feedback. We developed a detailed 15-20 minute survey comprising open-ended and qualitative questions, which we administered in local coffee shops and the Evanston Public Library. This approach enabled us to collect valuable, firsthand insights from the target audience.
  • Design Process: Based on the research findings, we utilized Figma for designing user-centric interfaces that address the identified issues. We created high-fidelity wireframes and interactive prototypes to visualize and iterate on design solutions. The prototyping phase was crucial in translating user feedback into tangible design elements, allowing us to test and refine our concepts effectively.
  • Product Testing and Evaluation: Following the design and prototyping stages, we conducted product testing sessions with participants from the older adult demographic. This iterative testing process provided critical feedback, which informed further refinements to our design. We rigorously evaluated the usability of the proposed solutions, ensuring that the final product aligns with the needs and preferences of the target users.
  • Key Findings and Design Implications: Our research and testing revealed significant challenges related to mobile interface accessibility and financial considerations of online shopping for older adults. These insights underscored the necessity for a more user-friendly design and enhanced support systems tailored to this demographic. This project underscored the importance of a consumer-centric design approach and the value of integrating real-world feedback into the development process.
  • Access the Full White Paper: Download the White Paper

Re-Designing Website for Cocktail Mixer Start-Up

As an analyst in Form and Function marketing, a student-run marketing agency, I worked on an account team for an Asian cocktail start-up. I was part of the website development team and produced weekly deliverables.

Brand Moodboard

I worked with project lead the website planning through taking in input from by our client and client teams. I created a themed guide for their overall aesthetic.

  • Details

Wireframe

The wireframe created direction for the website planning creating a clear framework for our sprints.

  • Details

Brand Values and Iconography

I consulted with the CEO of the company in order to understand the brand messages and then ideated iconography to match with the theme.

  • Details

Brand Content and Copy

I created three full copies with full imagery for inspiration for website. My work was used as stand in until new imagery was produced for Nomi.

  • Details

About Me Copy and Design

I created an interim about me page which included design and copy writing.

  • Details

Interpreting Medical Insurance Premium with Python Regression Modeling

During my data science rotation, I contributed to a project involving a medical insurance premium dataset from Kaggle.

  • Team Collaboration: Worked in a team of four to analyze and model the dataset.
  • Model Focus: Specialized in developing decision tree and random forest models to predict insurance premiums.
  • Data Considerations: The premiums were calculated based on demographic characteristics and past medical history.
  • Model Optimization: Used regression models since premiums are a dollar amount. Optimized the models using RMSE, which penalizes outliers and deviations from the line of best fit more than MAE, which treats all errors equally.

Leading Strategic Marketing Initiatives at Northwestern's Form & Function Marketing Consultancy

At Form & Function Marketing Consultancy in Evanston, IL, I have undertaken the following roles and responsibilities:

  • Led the Creation of 30 Strategic Presentations: Designed and delivered presentations on pre-professional topics and marketing strategies, including agile marketing, consumer intelligence, UX/UI, and product marketing, enhancing the skills of 60 general members.
  • Optimized SEO and Boosted Web Traffic by 225.7%: Conducted in-depth research and implemented SEO strategies, resulting in a significant increase in site traffic. Developed a 3-month content calendar for LinkedIn with 20 impactful posts, complete with alt-text and graphics, for an ed-tech start-up.
  • Redesigned Website Leading to 3.2% Conversion Rate: Overhauled the website for a cocktail mixer start-up using HTML and CSS, achieving a notable improvement in conversion rate.