Harshal Kamble

Built SQL-driven dashboards using Microsoft SQL Server and Power BI to resolve inconsistent KPI tracking across five rehabilitation programs. Integrated EMR data to analyze functional gains, length of stay, and discharge outcomes. Automated refreshes and advanced DAX reduced executive reporting time by 35%. Developed Python-based predictive and anomaly detection models, improving operational efficiency by 12% and reducing average length of stay by 2.3 days.

Optimized workforce analytics pipelines processing data from 150+ fulfillment centers. Rebuilt pipelines using advanced SQL (recursive CTEs, window functions, complex joins) and migrated to AWS Redshift with incremental loads. Used Python (pandas, SQLAlchemy) for validation, improving performance by 40%. Designed star schemas in Snowflake and integrated ARIMA/Prophet forecasts into Tableau with AWS Kinesis, enabling 4× daily updates and reducing decision lag by 80%.

Consolidated clinical data from 5+ siloed sources (EMR, labs, therapy, monitoring) using Python (pandas, pyodbc) and SQL ETL pipelines. Standardized ICD-10/CPT codes and built Azure SQL staging/production layers, improving data quality by 40% for 2,000+ patients. Built ML pipelines (scikit-learn, XGBoost) for cognitive fatigue prediction, improving accuracy from 60% to 81% and reducing incidents by 22%. Developed Power BI dashboards with advanced DAX, RLS, and automated refresh, saving 3 hours weekly.

Built automated Python pipelines using NLP libraries (NLTK, spaCy) for complaint text processing, sentiment analysis, and categorization. Integrated OpenAI API for response suggestions and implemented TF-IDF with cosine similarity, achieving 81% categorization accuracy and improving response alignment from 36% to 77%. Optimized SQL databases using stored procedures, indexed views, and query tuning; automated reporting via SQL Server Agent and SSIS, reducing resolution time by 35%.

Built centralized SSIS ETL pipelines integrating SAP ERP, MES, and regional databases across 80+ business units. Used Python for validation and transformation, and developed Power BI dashboards with drill-downs to standardize KPIs across 12 countries, improving resource allocation by 20% and enabling $2.3M budget reallocation. Analyzed sales data using advanced SQL and built ARIMA and Random Forest models to identify high-growth products, contributing $4.1M in additional revenue.