SHREEKANYA KODATE

Led a data-driven project that optimized the recommendation system by leveraging the k-prototypes clustering algorithm to gain valuable insights into customer behavior, demographics, and firmographics. Implemented the SPADE algorithm to establish cross-sell path, resulting in highly accurate and targeted recommendations, ultimately enhancing customer engagement. Created a dashboard in R-shiny. Mentored project on customer suspect prediction, utilizing ML models (XGBoost and ensembles). Developed, optimized, and maintain a robust cross-sell path dashboard, providing valuable insights to users. Successfully boosted product sales and decreased churn rate by leveraging an XGBoost model trained on feature-engineered customer data, enabling accurate sales opportunity prediction and facilitating data-driven decision-making to the leadership. Performed extensive Exploratory Data Analysis (EDA) and feature engineering to ensure high-quality data preprocessing. Developed and productionized time series model (SARIMAX) for revenue forecasting, providing valuable information for future revenue projections and enabling sales teams to make well-informed decisions. Exploring LangChains, and ChatGPT for potential implementation of copilot on our dashboard, showcasing a proactive approach to integrating advanced AI capabilities for enhanced functionality and user experience. Spearheading the upsell project and achieved a R2 score of 86% in experiments.

Supervised experiments and projects in Communications lab and Electronics Design lab.

Performed image segmentation on image using Otsu’s thresholding method from scratch and successfully extracted the first row and first letter from a text image. Achieved image enhancement by implementing histogram equalization.