Built an AI forecasting system achieving 18.35% accuracy across 100K+ e-commerce transactions. Predicts product demand 30-90 days ahead using Prophet models, helping businesses reduce overstock by ~23% and prevent stockouts. Key Features: ∙ Processes 100K+ transaction records ∙ Identifies seasonal patterns and trends ∙ Interactive dashboards for decision-making ∙ Optimizes inventory purchasing Tech Stack: Python | Prophet | Streamlit | Plotly Perfect for e-commerce brands managing complex inventory.