Omkar R. Email & Phone Number

Chicago, Illinois, United States

Greater Chicago Area

Digital Messaging, Customer service and engagement1. Behavior Analysis Dashboard Development - Designed and developed a dashboard for in-depth analysis of agent tenure, behavior, and performance metrics. 2. Stuck Conversations Tracking - Created a Tableau dashboard to monitor and resolve stuck messaging conversations promptly, resulting in a 5% reduction in monthly stuck conversations.3. Digital Messaging Volume Forecasting - Developed a predictive model using Python, fb-prophet, and SARIMA models to forecast incoming digital messaging volume. Achieved an RMSE of 7% for average monthly volume, optimizing agent staffing opportunities.4. Third Shift Agent Performance - Designed a dashboard to track Third Shift agent performance metrics, enhancing agent accountability.5. Agent Staffing Recommendations - Recommended staffing adjustments based on agent performance metrics, resulting in a 9-minute reduction in response times during peak hours.6. Performance Score Monitoring - Designed and developed a dashboard to monitor the impact of agents checking their performance scores on query resolution performance. This led to a 16% faster response time improvement over 6 months.7. Scorecard Dashboard Automation - Automated the implementation of a comprehensive scorecard dashboard in Tableau and SQL, featuring 23 metrics for daily monitoring and strategic implementation of Digital Messaging customer service and engagement. Implemented daily updates using Airflow and Python.8. CSAT Scores and Intent Analysis - Built a Tableau dashboard and conducted analyses to track CSAT scores and volumes of 180+ message intents handled by bot agents. This effort identified intents with the highest impact on customer experiences.9. Half-Hourly Staffing Prediction - Utilized agent availability and performance metrics to develop a staffing prediction model using H2O libraries in Python, achieving predictions with an MSE of 3 FTE per interval.

Atlanta

1. Share of Wallet Prediction for SMB Customers - Developed an advanced machine learning model to predict the share of wallet for Small and Medium Business (SMB) customers at UPS, resulting in significant improvements over earlier tree-based models.2. Churn Prediction for US-based SMBs - Conducted feature selection using Principal Component Analysis (PCA) and created a Logistic Regression model to predict churn among US-based SMB customers over the past three years. Achieved an impressive accuracy rate of 74% using Python and Google Big Query.3. Churn Prediction for Canadian SMB Customers - Built a Random Forest Classifier to predict churn for SMB customers in the Canadian region, achieving a commendable accuracy rate of 71%.

Atlanta, Georgia

1.Small Business Big Data Tool- Developed a sophisticated model to provide actionable recommendations for identifying new market entry points for Small and Medium Businesses (SMBs) at both the county and zip code levels.- Leveraged comprehensive data sources, including data from the Small Business Administration (SBA), Census, and the Internal Revenue Service (IRS), to inform the model.- Integrated the model seamlessly with a user-friendly Tableau-based dashboard, designed to address key questions for small business owners such as their current market position, the characteristics of their target population, and potential areas for business expansion.- Conducted in-depth analyses at the county and zip code levels, utilizing publicly available data from SBA, Census, and IRS, to create a powerful visualization tool.- This tool empowered SMBs to make informed decisions about their market expansion strategies, leveraging data-driven insights for growth and success.

Greater Atlanta Area

1.Employee/IC Classification Project- Conducted an analytics project funded by the Georgia Department of Labor, employing advanced analytics tools to analyze court opinions from the Free Law Project's RECAP archive and the federal courts' PACER system.- Utilized text mining techniques with R and Python and conducted content analysis to investigate various aspects, including plaintiffs' misclassification win/loss rates, legal tests employed by courts, and factors influencing judges' decisions.- Employed machine learning, regex, web scraping, and natural language processing (NLP) skills to process over 0.5 million documents, identifying key factors influencing judges' decisions and forecasting case outcomes.- Developed web scraping scripts in Python (BeautifulSoup and Scrapy) to extract court documents from Google Scholar.- Utilized regex for document searching, pattern matching, and filtering, contributing to the construction of the research corpus.- Collaborated closely with legal professionals and legal analytics research teams, with a focus on employment law.2.Ravel Caselaw Analysis Project- Conducted data extraction from Ravel Caselaw databases, which primarily involved working with unstructured data.

Mumbai Area, India

1.Consumer Market Knowledge - Iterative Forecast Tool- Collaborated with the Business Analyst (BA) for the US Baby Care sector to transform existing Excel-based forecasting sheets into a sophisticated, user-friendly tool.- Designed and developed an all-in-one forecast dashboard and simulator tool in TIBCO Spotfire, facilitating data forecasting and centralized access for multiple users.- Visualized crucial Key Performance Indicators (KPIs), including expenditures on Pricing, Promotion, Trade, and Media, along with their forecasted and actual values.- Empowered the BA to control official forecast values, while also creating a simulator capable of predicting volume changes based on adjustments to the four KPIs, one KPI at a time, for up to three months.2.Numerical Distribution Analysis – Indonesia Shampoo Traditional Channels- Addressed a critical business need by establishing a network to drive Numerical distribution, quantifying Key Business Drivers.- Utilized Bayesian Network analysis, employing a tree structure to examine variables influencing target variables based on Overall contribution, P-values, and Correlation.- Delivered a simulator capable of predicting numerical distribution, incorporating trained models at the trade sub-channel and regional levels. This utilized data from media, distribution, and sales fundamentals.3.Strategic Revenue Management - Revenue Pulse- Contributed to the development and implementation of the Revenue Pulse Project, a pivotal tool for assessing and optimizing pricing and promotional strategies for a Fortune 500 client in the Consumer Packaged Goods (CPG) domain.- Assumed responsibility for end-to-end implementation of the Spotfire-based solution, ensuring monthly deliverables, including ad hoc Excel reports, change request implementations, and data refreshes for backend tables.- Supported the business in determining upper and lower corridors for price index, aiming for maximum sustainable profitability.

Bachelor Of Engineering (Be), Electronics And Telecommunication

Master Of Science - Ms, Data Science And Analytics