I am a Master’s student pursuing Business Analytics. My coursework includes topics like Data management, statistics, predictive modeling and data visualization. I have worked on projects which range from simple dashboarding to advanced applied analytics. I have experience with tools and languages like SQL, Tableau, R and Python. I am looking to invest all the knowledge that I gained into a full time position and company where I can contribute to my fullest and learn new things at the same time.
-
Experience
• Business problem: To build and deploy forecast models to determine future customer demand for various products that is to be leveraged by the entire Centre of Excellence Department at Anixter
• Impact: Built a process which automatically forecasts future demand of over 90,000 products with a 30% increased accuracy and projected savings of up to $50 million per year
• Data cleaning: Extracted and cleaned 25 million records of relevant data using MS SQL by joining multiple tables and automated this process using stored procedures and views. Linked SQL tables to R console for cleaning and model building
• Modeling: Developed Machine Learning and Time Series forecasting models like Time Series Linear Model (TSLM), Holt’s, Moving Average, Seasonal ARIMA, Holt Winter’s and Neural network with Auto-Regressors for demand forecasting. Employed R packages like tidyverse, dplyr, ggplot2 and sweep for data manipulation and feature engineering
• Visualizations and Deployment: Developed a PowerBI interactive dashboard with forecasts, KPIs like error metrics and information of top 10 customers’ behavior at various granularities. Deployed the models to production using windows task scheduler with total project runtime of about 5 hours
-
Projects
DATA ANALYTICS PROJECTS
Grove Properties LLC – Analytics Capstone Project |Python, Tableau, ArcGIS|
• Identified an Investment portfolio worth $15 million in Charlotte with 2.7% higher growth rate than average market trend
• Executed Geo-spatial analysis to find hotspots based on multiple location features to visualize the best areas to invest
Voya Financial – Email Data Challenge |Python|
• Determined characteristics that drive success by Analyzing 2.6M subject lines of email data using nltk & textblob libraries
• Leveraged text analytics to identify and suggest the frequency and time to send emails for maximum client engagement
Kaggle – Big Mart - Sales prediction |Python, Tableau|
• Found features influencing sales and suggested ways to increase sales by performing Exploratory Data Analysis (EDA)
• Constructed a Linear Regression model to predict sales of different items with an Adj RSquare - 0.92 and low RMSE value and designed dashboards in Tableau to show sales distribution based on location and size of the mart