Arvind Chavan

Cash Forecasting for ATM Cash Management Company :

• Developed an ensemble model to predict the cash dispensed every day on ATM terminal level
• Analysis help to reduce the Cash outs and decrease the penalty imposed due to excess cash in ATM's at the end of each day
• End to end solution help to reduce manual processing of team
• Assessed accuracy and effectiveness of new and existing data sources and data analysis techniques.
• Collaboration with client and KPMG team to understand gaps in problems and communication.

Demand Sensing for US Pharmaceutical(for Ecommerce):

• Leverage AI/ ML to drive more accurate forecast by using various algorithmic models for each of customer code and product code combination and selection of the best model giving the lowest MAPE to facilitate production
• Collaboration with multiple teams to understand end to end process of data management
• Presentation of end results with multiple leaders within organization

Prediction of Repeat Caller for auto finance call centre :

• Understand the trends and patterns in data based on different KPI's and developed a statistical model to find important variable which are impacting repeat callers
• Interact with relationship manager to understand business which help to improve result
• Understand which call category is impacting more on repeat callers
• Focus on improvement of statistical model accuracy improvement

Classification of Senior Citizen Complaints :

• Data preparation based on business define rules for retail bank
• Data preparation and data cleaning using NLP data cleaning processes
• Experimented with various types of models — Random Forest, Decision Tree, KNN, Gradient Boosting, and XG Boost
• It help end user to reduce manual processing of complaints and improved overall rating of services

• Responsible for timely and accurate delivery of forecasted promoted and non-promoted volumes across various retailers in Canada
• Review and analyze historical sales trends, research demand drivers, prepare forecast data, develop statistical forecast models, and evaluate forecast results
• Compute volume forecasts at multiple levels of aggregation for multiple time horizons as part of a Canada Trade team
• Interact with multiple functions/teams to understand demand forecast drivers
• Utilize a collaborative and consensus approach by working with Trade and Customer Teams to obtain and ensure that current and accurate information is used for demand forecasts
• Effectively contributing to team initiatives for cycle time reduction by building automation modules for repetitive and cyclical tasks using tools such as R

Store level Demand forecast, Price Optimization, Promotion Effectiveness

• predict the female turnover in the United States for Eaton employees using statistical model.
• Interact with HR and leaders to find out exact solution after identification of high risk employees.
• Effectively provide solution to function HR.
• Ensuring high accuracy and quality of the deliverables by preparing and following quality checklist with all the important aspects for validating database .
• Identify purchase pattern of products, products tend to be purchased, analyze segment wise overlapping of products.
• Analyze different effective combination of product using statistical modelling.
• Preparation of user-friendly dashboard for customer.

Churn Prediction, Market Basket Analysis, Workforce Optimization