Developed dendograms using hierarchical clustering and item association rules in R to spot patterns in customer-item interactions, helping merchants determine popular product bundles. Performed text mining on key product attributes that drive customer behavior facilitating assortment decisions based on shopper preferences
Implemented k-means clustering algorithm in R to identify store clusters based on category performance and competitor threat. Developed 3 X 3 matrix framework, to visualize the cluster interactions. Principal component analysis was conducted prior to clustering to reduce the dimensionality of the dataset
Identified regional store clusters based on cross-category performance that helped the merchandising finance team to identify potential store-category combinations for diversifying the product offering. Principal component analysis was conducted prior to clustering to reduce the dimensionality of the dataset
Techniques Used
Hierarchical and k-means clustering, principal component analysis, market basket analysis, text mining
Developed a multivariate regression model in SPSS to quantify impact of a phased-out leadership development program on store employee turnover rate empowering talent development team to choose the right target audience
Developed linear mixed effect model to determine fixed and random effects that impacted merchant’s performance. Led efforts to map cross-functional operational metrics to HR data helping the compensation team revamp merchant recruitment and evaluation efforts
Techniques Used
Multivariate regression, linear mixed effect model, hypothesis testing, ROI determination
Carried out cost benefit analysis to make build versus buy decisions. Evaluated RFPs and streamlined the quote to pay process. Influenced build a data driven approach to vendor management and led the vendor management efforts
Conducted market research to study the latest trends in the industry and benchmarked the organization’s performance against the competitors. The recommendations were used to fund future product development initiatives
Built a binary classifier model to predict if the salary exceeded 50k based on the candidate profile. Overall, 82% prediction accuracy was achieved on the validation set.
As part of kaggle challenge, performed text data analysis and developed a multiclass classifier that predicted the type of cuisine based on the ingredients
Developed predictive model using decision trees in SAS miner to identify profile of students that are most likely to graduate and the supporting data exploration were in SAS visual analytics
Techniques Used
Decision Trees, SAS mining
Organizations and Causes
Co-Founder
Humans for AI non-profit, that aims to create a diverse workforce for the AI driven future
Lead Voulnteer
Asia Pacific Associate Network (APAN) resource group at Walmart
Student Volunteer
Women In Technology (WIT) group at University of Arkansas
