Mayur Shinde

I am a data scientist with 4+ years of experience building large-scale supply chain intelligence platforms at Hewlett Packard Enterprise. I specialize in advanced demand forecasting, deep learning, and production-grade inference pipelines that drive measurable improvements in planning, inventory, and fulfillment.

I led the design and deployment of transformer-based and ensemble forecasting architectures (ATTIRE & SHALYA), reducing MAPE from 60% to 35% and enabling inventory optimization valued at over $500M. I built resilience tools like SHAQUNI to quantify lead-time impacts and integrated them into priority frameworks to prevent bottlenecks.

I have delivered full-stack ML integrations into the o9 Digital Brain, created order-based ML forecasting and model-selection engines, and developed monitoring and alerting systems (NEWS) to surface early warning signals for stockouts. I also created conversational analytics interfaces using Streamlit and SpaCy to make forecast drivers accessible to planners.

I mentor cross-functional teams on scalable ML and Responsible AI, bridge research and enterprise deployment, and continuously iterate on systems to increase automation, planner trust, and operational impact. I hold a dual B.Tech + M.Tech in Industrial & Systems Engineering from IIT Kharagpur.