Supply Chain Analytics Expert with 15+ Years of Experience

Transforming complex supply chains through data-driven insights, simulation modeling, and intelligent automation. Currently driving spare parts optimization at Applied Materials using ML and discrete-event simulation.

Get in Touch Read My Articles

Key Achievements

$18M+
Cost Savings Delivered
20%
Cycle Time Reduction
15+
Years in Supply Chain

Professional Experience

Applied Materials, Maruti Suzuki, and more

Featured Projects

Simulation, optimization, and AI solutions

Latest Articles

Insights on supply chain, AI, and analytics

About Me

Supply Chain Analytics Expert | AI Enthusiast | Problem Solver

[Your Photo Here]

I'm a Supply Chain Analytics Expert with a unique blend of mechanical engineering foundation and advanced analytics expertise. Currently at Applied Materials, I drive innovation in spare parts planning using statistical forecasting, discrete-event simulation, and machine learning.

My journey spans from transmission design at Maruti Suzuki to leading global logistics optimization initiatives. I specialize in translating complex operational challenges into data-driven solutions that deliver measurable impact.

Beyond traditional analytics, I'm passionate about agentic AI and its potential to revolutionize business process automation.

Skills & Expertise

Analytics & Visualization

Python SQL Tableau Hadoop Databricks

Supply Chain Tools

Supply Chain Guru AnyLogic Celonis ARIS

Optimization & Methods

Gurobi Linear Programming Mixed Integer Programming Discrete Event Simulation

AI & Automation

MS Copilot Studio Claude Code Process Automation Machine Learning

Professional Experience

15+ years driving supply chain excellence across semiconductor and automotive industries

Mar 2024 — Present
Applied Materials

Spares Planning Analyst

Applied Global Services — After-sales service and spare parts lifecycle management

  • Lead business improvement projects using Machine Learning and Simulation
  • Conduct scenario planning using End-to-End Supply Chain Simulation Model
  • Develop warehouse capacity forecasts using S&OP forecast and growth trends
Oct 2017 — Mar 2024
Applied Materials

Logistics Network Optimization Analyst

Applied Global Services — Global logistics and network optimization

  • Achieved ~10% cost savings and 20% cycle time reduction through automated reverse logistics
  • Led tariff mitigation project saving $5M+ in duties (AGS Quarterly Team Award 2019)
  • Saved $2.4M through Cartonization Algorithm for warehouse outbound packaging
Jul 2008 — Jul 2015
Maruti Suzuki India Ltd

Transmission Design & Supply Chain Localization

R&D Division and Supply Chain Division

  • Delivered $10M in savings through localization price negotiations
  • Managed transmission parts development from design to mass production

Projects

Simulation models, optimization tools, and AI-powered solutions

Spare Parts Supply Chain Digital Twin

Built end-to-end discrete-event simulation models in AnyLogic to replicate the global spare parts supply chain for semiconductor capital equipment. The model captures stochastic lead times, multi-echelon inventory policies, and capacity constraints across regional depots to run what-if scenarios for disruption response, safety stock calibration, and capacity planning during demand spikes.

AnyLogic Simulation Digital Twin

Agentic AI for Supply Chain Workflows

Designed autonomous AI agents using Copilot Studio and Claude Code to automate repetitive supply chain workflows -- from demand signal processing and exception-based routing decisions to automated report generation and supplier communication drafts. These agents operate within defined guardrails, compressing the detect-decide-act loop from hours to seconds while keeping a human in the loop for high-stakes decisions.

Agentic AI Copilot Studio Automation

ML-Based Spare Parts Demand Forecasting

Developed machine learning forecasting models for high-value, low-volume semiconductor spare parts where demand is inherently intermittent and lumpy. Benchmarked traditional methods like Croston's and SBA against gradient-boosted trees and neural network approaches, ultimately deploying an ensemble model that reduced forecast error by 30%+ over legacy statistical baselines while handling the heavy zero-demand periods that break standard time-series methods.

Machine Learning Forecasting Python

Global Reverse Logistics Network Optimization

Formulated and solved mixed-integer programming models in Gurobi to optimize the global reverse logistics network for semiconductor equipment returns and refurbishment. The model incorporated duty/tariff structures, cartonization algorithms, and depot-to-customer mapping to minimize total landed cost. Delivered $2.4M in annual savings and a 20% reduction in reverse logistics cycle time through optimized routing and consolidation strategies.

Gurobi Optimization Reverse Logistics

Articles

Thoughts on supply chain, AI, simulation, and analytics

Supply Chain Mar 2026

Digital Twins Are Not Dashboards: Using DES as an Operational Planning Engine

Most "digital twin" implementations in supply chain are glorified monitoring layers. A true operational digital twin, built on discrete-event simulation, lets you stress-test your network against scenarios that haven't happened yet -- port closures, demand surges, supplier failures. In semiconductor supply chains, where a single tool-down event can cost a fab millions per day, the ability to run 500 disruption scenarios overnight and pre-position inventory accordingly is not a nice-to-have. It is a competitive requirement.

8 min read
AI & Automation Feb 2026

Agentic AI Beyond Chatbots: When Supply Chain Systems Start Acting, Not Just Advising

The real shift in supply chain AI is not better predictions -- it is autonomous execution. Agentic AI systems can now monitor demand signals, trigger replenishment orders, reroute shipments around disruptions, and draft supplier negotiations without waiting for a human to click "approve" on a dashboard. The key challenge is designing the right guardrails: which decisions can an agent own outright, which need human oversight, and how do you audit an AI that made 10,000 routing decisions overnight?

12 min read
Analytics Jan 2026

The Intermittent Demand Problem: Why Standard Forecasting Fails for Semiconductor Spares

When 80% of your demand history is zeros and a single order can be worth $50K, conventional time-series forecasting breaks down completely. Croston's method and its SBA variant were designed for this, but they assume stationary demand patterns that rarely hold in practice. Modern ML approaches -- gradient-boosted trees that incorporate installed base data, equipment age, and usage telemetry -- can outperform statistical baselines significantly. The real gains, though, come from ensemble methods that blend classical intermittent-demand models with ML regressors.

10 min read

Insights & Notes

Quick thoughts, observations, and learnings

Note Mar 2026

Why Simulation Exposes What Spreadsheets Hide

After years of watching spreadsheet-based planning models fail under pressure, the pattern is clear: static models cannot capture lead time variability, capacity constraints during demand spikes, or the bullwhip amplification that cascades through multi-echelon networks. A simulation model we built revealed that a 15% increase in demand variability at the customer level was creating a 45% inventory swing at the regional depot -- a dynamic completely invisible in the monthly planning spreadsheet everyone had been trusting for years.

Observation Feb 2026

AI Is Moving from Prediction to Action -- and Operations Teams Need to Be Ready

The most significant shift in supply chain technology right now is not better algorithms -- it is the move from AI that recommends to AI that executes. Agentic systems are already handling millions of routing and replenishment decisions at leading logistics firms. For operations professionals, this means the job is evolving from making decisions to designing the decision frameworks and guardrails that autonomous agents operate within. The planners who thrive will be the ones who learn to manage AI agents, not compete with them.

Resources

Curated tools, books, and references

Recommended Books

Factory Physics -- Hopp & Spearman

This is the book I wish someone had handed me in my first year of operations work. Hopp and Spearman replace intuition and tribal knowledge with rigorous, physics-based laws governing throughput, inventory, and cycle time. If you want to understand why your supply chain behaves the way it does -- not just what to do about it -- this is the foundation. The variability buffering framework alone changed how I think about safety stock.

Supply ChainOperations

Inventory and Production Management in Supply Chains -- Silver, Pyke & Thomas

The definitive reference for anyone serious about inventory optimization. Silver, Pyke, and Thomas cover every major inventory model with mathematical rigor but also practical context -- from continuous review (s,Q) policies to periodic review (R,S,s) systems. I still reference this book when designing inventory policies for spare parts. It bridges theory and implementation better than any other text in the field.

AnalyticsInventory

Useful Tools

AnyLogic

My go-to simulation platform because it is the only tool that natively supports discrete-event, agent-based, and system dynamics modeling in one environment. For supply chain work, this matters -- you need DES for warehouse and order flow modeling, agent-based for supplier behavior, and system dynamics for strategic capacity planning. The ability to embed custom Java logic, connect to live databases, and deploy models as interactive web apps makes it practical for both analysis and stakeholder communication.

SimulationDigital Twin

Gurobi Optimizer

When a supply chain problem can be formulated as a mixed-integer program -- network design, vehicle routing, facility location, inventory allocation -- Gurobi is the engine I reach for. It consistently solves large-scale MIP models faster than alternatives, which matters when you are running sensitivity analyses across hundreds of tariff scenarios or optimizing reverse logistics routes across a global network. The Python API (gurobipy) integrates cleanly into data science workflows, making it straightforward to go from analysis to production.

OptimizationOperations Research

Education

Academic background and certifications

Jul 2015 — May 2017

Master of Science in Industrial Engineering

Oklahoma State University, Stillwater, USA

Jul 2004 — May 2008

Bachelor of Technology in Mechanical Engineering

G B Pant University, Uttarakhand, India

Get in Touch

Open to opportunities in supply chain analytics, AI implementation, and operations optimization

Email

rajeev.gangwar@hotmail.com

LinkedIn

/in/gangwarrajeev

Phone

+1 (405) 334-8592

Send a Message