Data Analyst · Applied Data Scientist · Python · SQL · Power BI · AWS
Hands-on experience across the full analytics lifecycle data cleaning, feature engineering, time series forecasting, dashboarding, and predictive modeling. Background in engineering and data quality brings a strong focus on data integrity, reproducibility, and reliable decision-making across retail, operations, and research environments.
End-to-end AI compliance assistant using Python, Flask and LLM API to automate audit and compliance queries for 800+ real business clients — reducing response time from hours to under 3 seconds with 24/7 availability. Engineered document intelligence with pdfplumber for instant structured summaries from government notices. Identified and remediated 6 critical vulnerabilities including cross-account session hijacking and 17 categories of prompt injection attacks before production deployment.
EDA on supply chain data including sales, inventory, transportation, and defect rates. Analyzed revenue by product type and SKU, transportation costs, defect rates across carriers and shipping modes, plus inventory metrics such as stock levels, order quantities, and lead times.
Interactive Power BI dashboard analyzing 7,120 production records across 8 machines and 3 shifts. Identified 69.62% OEE vs 85% benchmark. Assembly Line B emerged as the critical underperformer at 50.87% with 2,083 hours lost annually — nearly 4x the best-performing machine.
Predicts car insurance purchase likelihood using customer demographics, vehicle attributes, and behavioral patterns. Compared Decision Tree, KNN, and Random Forest classifiers evaluated using Precision, Recall, F1-score, and ROC-AUC to identify high-likelihood customer segments.
Personalized book recommendation system using collaborative filtering in Python. Deployed as a full end-to-end web application using Flask and Heroku, demonstrating practical model integration and delivery.
Automated categorization pipeline using NLP techniques and BERT-based text representations for scalable classification of unstructured text. Improved classification quality through advanced EDA, cleaning, and tokenization ensuring high-integrity model inputs.
Interactive profitability dashboards in Tableau analyzing product-, state-, and category-level sales and margin performance. Created P&L reporting views to identify revenue trends, margin drivers, and customer-level profitability insights.
Analytics tool processing unstructured WhatsApp chat exports to deliver automated insights into user activity, engagement, and conversation patterns. Features a user-facing web interface for uploading and analyzing chat data.
Open to Data Analyst and Applied Data Scientist roles. I bring a full-stack analytics mindset — from raw data to business decisions. Based in Mumbai, open to remote and relocation opportunities.