Anshuman Das

SKU Recommendation Engine:

The project involves generating Product level(SKU)recommendations for retailers and optimizing assortment sizes for the Sales team for efficient Retail Distribution and assisting the marketing team to facilitate better Category and portfolio management by measuring the market performance at SKU level.

• Collaborated with internal stakeholders, identifying and gathering analytical requirements for customer, product and projects needs.
• Engineering 29 variables from raw Point of Sales to improve model performance.
• Automated and monitored more than 14 Machine Learning Algorithms on quarterly basis using training dataset.
• Leveraged Statistical techniques to check data drift in pipeline to maintain performance of predictive models.
• Improved recall value and precision and recall value by 15% and 10% respectively hence bridging gap between relevant and recommended.

Data Scientist:
End Of Day Pricing:
Time Series Forecasting end of day close prices for financial products for efficient portfolio management by leveraging Risk and Return tradeoff.

• Implemented Time Series forecasting models like ARIMA,SARIMA for predicting chosen Price of asset.
• Calculate KPIs like volatility and returns for each asset for efficient portfolio construction with minimum risk and maximum returns.
• Using Statistical and model diagnostics for model evaluation.
• Evaluating complete ETL pipeline starting from data normalization to data finalization by comparing source with target at each staging step.
• ROI upon Portfolio optimization improved by 8.5% in financial year.