About this Presentation
1. The session challenges traditional buffer sizing assumptions, showing why static targets and backward-looking consumption data fail in high-variability distribution environments. 2. It reveals how AI/ML models can extract demand signals from noise at granular SKU-node levels, enabling more responsive and accurate demand prediction during replenishment lead time. 3. The presentation hints at how separating predicted demand, demand safety buffers, and supply safety buffers creates dynamic protection against both demand surges and supply instability. 4. It illustrates how dynamic buffer targeting improves availability, freshness, and assortment simultaneously—without increasing inventory or operational complexity.
Buffer targets in TOC implementations use past patterns of consumption data and are often inadequate to protect availability against demand shifts caused by predictable demand drivers. The proposed dynamic buffer targeting, using AI/ML in demand sensing, anticipates these demand shifts and adjusts the relevant buffers proactively, thereby reducing stockouts and excess inventory. Its implementation in multiple companies has produced encouraging results.
What Will You Learn
To help you get the most value from this session, we’ve highlighted a few key points. These takeaways capture the main ideas and practical insights from the presentation, making it easier for you to review, reflect, and apply what you’ve learned.
Instructor(s)
Dr. Rakesh Sinha
Dr. Sinha is the founder and CEO of Reflexive Supply Chain Solutions, a specialized consulting firm in the area of Operations and Supply Chain.
He has worked with Godrej Consumer Products as the Global Head of Manufacturing, Supply Chain and IT. He led the TOC implementation in GCPL in 2004, which was the first Viable Vision implementation in the world. Under his leadership, GCPL was awarded the Platinum Award by TOCICO in 2015. Dr. Sinha has led several TOC implementations across India, USA, Indonesia, Africa and Latin America.