About

Adarsh is a customer-focused analytics professional with 5+ years of experience in the supply chain domain using decision science, machine learning, and statistical algorithms. He has conceptualized and implemented multiple strategic and transformation solutions in the demand planning, distribution planning, and regional planning space. Adarsh has helped businesses navigate complexities by enabling collaboration among diverse stakeholders across multiple geographies. He has driven consumption by storyboarding insights/recommendations to business leaders and enabled data-driven decision making. Adarsh has developed the statistical demand forecasting process for the entire Cisco product offerings and for the entire AWS spares to support 7.5M hosts in 90+ clusters. He has also designed the inventory balancing engine to have an optimum inventory level of AWS spares at site level across 23 countries and 30+ clusters. Currently, Adarsh works as a data scientist at Amazon, where he has modelled the spider ensemble forecast for generating a time series statistical forecast for 1000+ AWS spares to improve forecast accuracy from 45% in 2020 to 62% in 2021 and part availability to support 7.5M hosts in 90+ clusters for 1200+ AWS spare parts. Adarsh has implemented room inventory planning for centralized warehouse clusters to improve part availability at site to 93% in Aug 2021 from 85% in Feb 2021 by proactively identifying the spares demand using text analytics and web scraping of trouble tickets/SIM tickets raised by DCO team. He has also designed the inventory balancing engine using PuLP package for 23 countries, 30+ clusters, 500+ parts to balance the excess and shortage inventory at different sites worth ~$10.5M driven by business constraints to have an optimum inventory level at each site. Adarsh has worked as an integrated business planning manager, distribution planning, sales planning, global planning, and summer intern at Cisco. During his tenure, he spearheaded a team to predict ageing inventory at CM sites, hubs, and distis to lower financial risk resulting from it. He has devised a booking adherence scorecard metric for inventory planning to measure co-planners’ accuracy and to help guide distis on the units and corresponding value to be booked for the week. Adarsh has also implemented an MVP to forecast for products’ Disti booking looking at historical Disti POS data, end customer demand, current pipeline status, lead time, safety stock