• Examined $1.5 mil. of sales data of a recently-acquired activewear firm to provide the new owner with insight on current/historical product lines with highest sales and regions with greatest concentration of sales, to streamline future product offerings/development and for region-/demographics-specific marketing
• Usage of NLTK for processing and categorizing 5000 unique product names
• Forecasted future demand of product categories given discount rates using GluonTS (Neural Nets), Prophet, & AutoTS
  IN PROGRESS:
• Developing a CI/CD pipeline and dashboard web-app with Plotly-Dash/Atoti & Heroku for forecasting & profitability

Github