● Extracted 'Item 1a Risk Factors' from 10-K documents using the SEC EDGAR API, facilitating financial risk analysis.● Designed and implemented a Spark Session leveraging PySpark for parallel text processing, reducing data processing time by 70% for the large dataset of 10-K documents.● Developed a comprehensive preprocessing pipeline using SparkNLP and calculated uncertainty scores for financial documents through an unsupervised learning approach, providing more precise risk assessments over CFIUS reviews.

I contribute to the tech team at IU Surplus, specializing in building and refurbishing computers and laptops, conducting repairs, and overseeing inventory management for the store.

● Led a team of 4 and developed data pipelines to consolidate disparate addresses into a unified data frame of 16,000 addresses, raising the count by approximately 33%.● Integrated additional geographic information into the data frame by leveraging Google’s API to enhance the dataset’s value that facilitated spatial analysis by integrating 3 new features.● Formulated a targeted business problem statement to optimize logistics operations and support a faster delivery model; through data analysis and process optimization, that led to a 15% improvement in delivery efficiency.● Integrated the application of Travelling Salesman Problem with K-means clustering to generate clusters revealing underlying patterns and spatial relationships, reducing the routing distance withing the clusters by 22%.● Engineered a comprehensive dashboard using Folium and Streamlit to visualize spatial data, including KPIs and interactive filtering; optimized decision-making processes and doubled efficiency in interpreting location-based insights.