Data Analyst with hands-on experience in supply chain analytics, ETL pipeline development, and statistical modeling. Skilled in Python, SQL, and BI tools (Power BI, Tableau), with a strong foundation in mathematical optimization, data mining, and forecasting. Leveraging data-driven insights to optimize logistics, inventory management, and business operations. Improving supply chain efficiency through predictive analytics and process automation.
Atharv IT Solutions
Managed data accuracy and compliance for operational databases, ensuring 99% accuracy in master data governance.
Conducted predictive analytics using machine learning algorithms to forecast power usage trends, reducing operational waste by 24% and optimizing energy allocation strategies.
Directed automated ETL processes using Python and SQL to transform and load large datasets into a centralized data warehouse, enabling real-time analytics and data-driven decision-making.
Created Power BI dashboards for tracking inventory performance and logistics, improving visibility into supply chain KPIs.
Atharv IT Solutions
Investigated and optimized data-driven marketing campaigns using Google Analytics, Excel, and CRM tools.
Produced a classification model to identify behavioral patterns, improving targeted interventions and reducing behavioral issues by 23%.
Created predictive analytics models for sentiment analysis, supporting strategic business decisions.
Designed interactive dashboards to track intervention success metrics, enabling real-time insights for psychologists and school administrators.
Stevens Institute of Technology
Mumbai University
Conducted NLP-driven analysis of 13000+ Airbnb posts across 4 cities to examine host sentiment and uncertainty (2014–2020). Achieved 85%+ classification accuracy using spaCy, Text Blob, and regex for uncertainty and sentiment detection. Discovered 34–51% of posts contained regulatory uncertainty, informing policy.
Built a fully automated trading pipeline using Python & Robinhood API, integrating real-time market data and execute trades. Engineered preprocessing and watchlist modules using volume reduce manual trading by over 90%.
Cleaned & transformed multi-month dataset for 10,000+ customers using imputation, outlier handling and Z-score. Collaborated and prepare a predictive pipeline using Random Forest, XG Boost, KNN and Decision Tree Models with Random Forests delivering top F1 Score. Pioneered visualization tools to track training history and model performance, improving interpretability.