Mihir Patel

Problem Statement: To forecast Sales using past Sales data of three years, corresponding Weather and Promotion Data for all product categories across four regions (as defined by client) of the United States, for planning and making appropriate business decisions • Client’s Industry Domain: Manufacturing • Solution Provided: The client’s problem statement was solved using Time Series Modeling. The technique used was Facebook’s Prophet Model. Forecast accuracy was reported using metric: MAPE, which were within reasonable and acceptable ranges as per the client. • Deliverables: o Best model (as per hyperparameter tuning and cross validation) for each Category across all regions and predictions for next quarter. o A Tableau Dashboard, consisting of consolidated plots indicating future potential Sales. The intention of creating the dashboard is to provide comprehensive visualizations for the client for appropriate decision-making regarding inventory management, production targets, planning of promotions, etc.

• Responsible for providing recruitment solutions to clients belonging primarily to the West (and often times South) zone of India • Managed mega interview drives for clients (including giants like Paytm, BharatPe) and led a team of around 11 members. • Used basic descriptive statistics to recognize nature of candidates and clients (like feedback TAT, intake, retention ratios, etc.) to prioritize, finalize target numbers for my team and accordingly allocated tasks among them as per available bandwidth and their abilities. • Redefined processes and eliminated redundant parts. Result was enhancement in operational efficiency and better conversion in consequent interview drives.

Relevant coursework: Simulation Modeling, Statistical Methods, Statistical Modeling, Data Visualization, Data Management, Time Series Forecasting, Data Mining, Data Wrangling, Data Analysis, Business Intelligence, Optimization