McLean, Virginia, United States
Retail Bank - Customer Protection
Client: Leading Hospitality and Entertainment Business 1. Optimized marketing budget allocation by 18% by identifying profitable promotions for 20 customer segments using causal analytics techniques (Fixed Effect Panel Data Regression, Propensity Score Matching) in Python and R. 2. Accelerated decision-making process related to promotional strategies and resource allocation by 25% with Tableau dashboards to communicate analytical findings and methodologies to cross-functional stakeholders. 3. Enhanced data consistency by 30% for over 3M records by spearheading the development of ETL processes using MySQL, constructing and maintaining a robust database architecture. 4. Enhanced project efficiency by 20% as a product owner by optimizing Jira agile workflows, product backlogs, and sprints. Client: Local Non-Profit and Community Development Financial Institutions 1. Enhanced retention by 18% with customized strategies by identifying 3 key churn predictors with detection boundaries leveraging Random Forest model and Logistic Regression in Python. 2. Slashed financial support project costs by 35% through detailed clientele analysis and customized planning recommendations via Exploratory Data Analysis (EDA) on 8 years of clients' data. 3. Identified key customer segments utilizing k-prototype clustering in Python, informing tailored CRM strategies. 4. Increased 13% customer satisfaction rate by developing an interactive Tableau dashboard to display financial improvement. Project: Walmart’s Retail Sales Volume Prediction Predicting sales volume of 28 days based on historical dataset spanning 5 years. Achieved a top 10% ranking with a 0.76 Root Mean Squared Scaled Error using LightGBM, LSTM, and Seq2Seq models with Python.
1. Gained a 10%+ return within 6 months in back-testing by developing sectoral rotation trading strategies based on Time Series and Neural Network algorithms. 2. Designed investment portfolio of Taiping Fund products based on risk parity and portfolio optimization principles, and back-tested the portfolio's performance.
• Expanded company database by 800K+ natural language data points using Python, resulting in a 20% increase in proficiency. • Conducted in-depth research on supply chain finance industry to enhance database augmentation effectiveness. • Achieved a 15% reduction in training time through data engineering efforts at Linklogis.