– Developed and fine-tuned multiple deep learning models for Time Series Forecasting (N-Beats, DeepAR) achieving over 93% accuracy in predicting sales and demand, decreasing overstocking by over 70%.– Developed an Unsupervised learning model (Kmeans++) to cluster products and define product-shelf placements increase product picking efficiency by 35% and overall packing times by 20%.– Innovated a user-friendly Python and Tabula application, automating incoming sales orders and reducing… Show more – Developed and fine-tuned multiple deep learning models for Time Series Forecasting (N-Beats, DeepAR) achieving over 93% accuracy in predicting sales and demand, decreasing overstocking by over 70%.– Developed an Unsupervised learning model (Kmeans++) to cluster products and define product-shelf placements increase product picking efficiency by 35% and overall packing times by 20%.– Innovated a user-friendly Python and Tabula application, automating incoming sales orders and reducing order processing time by 60%, thereby boosting operational efficiency.– Seamlessly deployed sophisticated deep learning models to production environment and created continuous improvement/Development CI/CD pipelines to maintain/improve prediction accuracy. – Conducted thorough analysis and cleaning of complex datasets, revealing key insights into sales and demand trends. Led the creation of robust data pipelines from diverse sources, enhancing decision-making accuracy. Show less