Building data products end-to-end
I'm a graduate student from University of Memphis, specializing in transforming messy data into structured & powerful insights through predictive modeling, statistical methods, and clean pipelines.
Driven by curiosity, powered by data. My passion lies in uncovering hidden patterns and trends that drive meaningful business decisions. From data cleaning to advanced analytics, I bridge the gap between raw information and actionable intelligence.
With a strong foundation in both technical skills and strategic thinking, I bring a unique perspective to every project ensuring data integrity while delivering insights that matter.
Outside of work, I enjoy chess and anime.
Specialized expertise in transforming complex data challenges into strategic opportunities
Ensuring data integrity from the ground up.
Forecasting outcomes with algorithms.
Storytelling through numbers.
Bridging insights to action.
Real-world applications of data science and machine learning
A stacking ensemble hybrid model to detect difference between right & wrong AI-generated text combining classical ML, a fine-tuned Transformer, and embedding dissimilarity.
Fine-tuned `distilbert-base-uncased` for deep semantic context
`StratifiedKFold` cross-validation for robust generalization
`LogisticRegression` meta-learner blends base predictions
Built comprehensive risk assessment for a P2P lending platform using advanced ML techniques.
Cleaned and preprocessed large-scale loan data using Excel & Python
Applied sophisticated imputation for missing data handling
Extensive EDA to identify key borrower risk features
Gradient Boosting model optimized for precision/recall
Delivered actionable risk insights and recommendations
Developed a logistic regression model to predict heart disease risk with comprehensive data analysis and visualization.
Led data cleaning in Excel with focus on handling class imbalance
Implemented feature selection techniques for optimal model performance
Created interactive Tableau visualizations (smoking, age, BMI, activity)
Selected final model based on rigorous accuracy metrics
Built a complete BEV in MATLAB/Simulink — powertrain, controls, and energy analytics — to study speed tracking, range, current flow, and regenerative braking behavior across a standard drive cycle.
Assembled RWD drivetrain: Physical Vehicle Body, DC motor, simple gear, and four tire models.
Implemented H-Bridge inverter driven by PWM; longitudinal PI driver for speed tracking (FTP75).
Battery & SOC pipeline: 240 V pack, 80 Ah; discrete integration for SOC with regen capture.
Results: ~16.5 km distance on FTP75 with ~15% SOC remaining; regen visible in current/SOC traces.
Documented component sizing & parameters (e.g., 75 kW DC motor, gear & tire coefficients).
From engineering foundations to analytics excellence
A commitment to continuous learning and hands-on experience