Financial Chatbot
AI Chatbot for Stock Market Analysis
Created an intelligent chatbot that combines large language models with financial data APIs to generate real-time stock insights, enabling users to make better investment choices.
Prediction Dashboard
Uber Driver Prediction Dashboard
Datathon 2025 Competition 🏆 “Best Use of StrataScratch DC” 🏆
As part of Team SYNTAX ERROR, developed a machine learning dashboard to predict whether new Uber sign-ups would convert into active drivers. Combined XGBoost with KMeans to analyze onboarding patterns, achieving 0.97 AUC. SHAP analysis highlighted onboarding delays and signup source as key drivers.
Review Sentiment App
Airbnb Sentiment Application & Hosting Strategy
Developed a sentiment analysis app for Airbnb hosts using NLP and XGBoost. The app predicts whether a review is positive or negative, visualizes key terms through word clouds, and allows users to flag incorrect predictions. These user-labeled responses are designed to support future supervised model retraining for improved accuracy and adaptability.
E-Commerce Database
E-Commerce SQL & Relational Database Design
Designed a relational data model and normalized schema for a global fashion retailer. Wrote SQL queries to support churn prediction, marketing attribution, return behavior analysis, and inventory optimization.
Anemia Prediction
Predicting Anemia with Machine Learning
Developed classification models using blood test data to predict anemia, focusing on early detection support. Leveraged feature selection techniques to identify the most relevant clinical indicators and applied SMOTE to address class imbalance. Models such as Decision Trees and Logistic Regression were used to explore interpretability and predictive performance.
Obesity Risk Analysis
Obesity Risk Analysis and Behavioral Clustering
Conducted a comprehensive analysis of obesity risk using classification models, clustering techniques, and ANOVA. Explored lifestyle and dietary factors to uncover patterns linked to obesity. After segmenting individuals into behavioral clusters, we provided tailored recommendations for each group to support targeted public health interventions.
Technology Consulting
Smart Grocery Shopping Assistant
Hands-on consulting course focused on applying AI and analytics to real-world business problems. Topics include use case development, prompt engineering, product management, Agile delivery, and stakeholder alignment. Culminates in a Shark Tank–style team project judged by industry experts.
Marketing Strategy
Repositioning Google Pixel — A Strategic Marketing Analysis
Analyzed Google Pixel’s weak market penetration despite technical strengths in AI and camera performance. Assessed brand positioning, competitive dynamics, and proposed strategic recommendations to enhance Pixel 9’s presence in the premium smartphone segment.
Explore Capstone Project
Johnson & Johnson Cerenovus Capstone Project
Developed a full-stack analytics dashboard to support sales and retention strategy at Cerenovus, a Johnson & Johnson medical device division. Integrated SQL-based data pipelines, machine learning models (including churn prediction and clustering), and interactive visualizations to identify key revenue drivers, at-risk accounts, and regional product trends. The tool enables real-time exploration and supports strategic planning across sales teams.
© 2025 Yiwei Lu. All rights reserved.