Krishna Mineshkumar Desai

• Architected and optimized an Entity Name Matching model to match incoming records against a 13M+ ground truth dataset, initially using PandasEntityMatching (XGBoost-powered) and later migrating to SparkEntityMatching for scalable, distributed processing. • Engineered entity resolution logic using Soundex, Double Metaphone, and other features to match against a 13M+ ground truth dataset, leveraging cosine similarity and Sorted Neighborhood Indexing (SNI) to generate high-quality candidate pairs efficiently. • Enhanced the core data-matching engine powering Project 990, significantly increasing match precision for grant seekers and grant givers and improving the reliability of downstream analytics and reporting.

• Refactored a large-scale extraction pipeline for 2M+ nonprofit records, building a scalable Python-based migration framework to convert unstructured IRS Form 990 filings into structured, queryable datasets. • Discovered and resolved thousands of critical duplicate records during pipeline refactoring, escalating the issue to leadership and implementing robust deduplication logic that improved data integrity and downstream model accuracy. • Built the “U.S. Education Philanthropy Desert” from scratch by preprocessing millions of records, performing advanced EDA, and developing layered geospatial visualizations in Tableau with dynamic filters and drill-down capabilities.

• Served as a Student Coordinator under Professor Namdeo Badhe for three years, managing placement activities such as interview rounds, training sessions, and pre-placement talks. • Coordinated with company HRs, scheduled placement drives, and maintained key records to support smooth execution. • Led communication between students, faculty, and recruiters, strengthening organizational and interpersonal skills. • Mentored junior team members and contributed to improving student preparedness through workshops and mock sessions.

• Analyzed 4 datasets and developed machine learning models for prediction and classification tasks. • Achieved model performance scores ranging from 0.92 to 0.99 across multiple use cases. • Built automated data workflows and integrated data from multiple sources for analysis and reporting. • Presented analytical findings through data visualizations to support business decision-making.

• Utilized Python and machine learning models to predict employee retention and support data-driven decision-making. • Designed and implemented data pipelines with ETL processes and standardization for data accuracy and reliability. • Created actionable insights and visualizations to effectively communicate findings to stakeholders. • Aligned analysis with key business performance metrics for strategic impact. • Developed predictive analytics solutions to optimize resource allocation and improve operational efficiency.

Coursework: Applied Machine Learning, Data Mining, Applied Database Technologies, Advanced Database Concepts, Applied Algorithms, Management, Access and Use of Big, Complex Data, Information Visualization, Data Visualization, Introduction to Statistics, Language Modeling

Relevant Coursework: Machine Learning, Data Mining, Mathematics (1,2,3,4), Soft Computing, Database Management System, Big Data Analytics, NoSQL