Supply Chain & Logistics · Operations · Data Analysis · Applied AI

I am a supply chain and logistics professional with 20+ years of field experience in operational environments, including selective retail, luxury, food logistics, warehouse operations, inventory management, and logistics coordination.

My professional background is mainly focused on:

• flow and stock management
• inventory reliability
• operational performance
• KPI monitoring
• SAP, WMS and ERP environments
• team coordination and field management
• logistics organization and process improvement

Today, I am strengthening this operational experience with data analysis, automation, and applied AI.

My objective is simple: use data and AI to improve operational processes, support decision-making, and help business teams work more efficiently.

My experience includes operational and project roles in logistics and supply chain environments.

Logistics organization, SAP/WMS improvement, inventory reliability, operational coordination, logistics budgets, and cross-functional optimization.

Warehouse team management, productivity and quality follow-up, picking optimization, inventory management, planning, and coordination with sales administration teams.

Stock management, supplier coordination, order follow-up, retail logistics execution, and operational control.

I have also worked in higher education in China, where I developed specialized training resources for engineering and supply chain programs. This experience helped me strengthen my ability to structure knowledge, explain complex topics, and design practical learning tools.

This GitHub portfolio documents my transition toward a stronger use of data, automation, and AI in operational contexts.

It is not a collection of theoretical AI experiments.

The objective is to build projects that are:

• practical
• understandable
• business-oriented
• connected to real operational problems
• useful for reporting, monitoring, classification, alerting, or decision support

I am particularly interested in how AI can be used inside companies to improve existing processes rather than replace operational expertise.

The projects in this portfolio focus on practical use cases such as:

• supply chain monitoring
• inventory and KPI reporting
• operational alerts
• process automation
• text classification for business workflows
• decision-support tools
• internal copilots
• local RAG systems
• AI-assisted analysis of operational data

The common idea behind these projects is to connect technical tools with real business needs.

Automatic classification of customer complaints using a neural network trained on more than 300,000 real-world texts from the CFPB dataset.

• Weighted F1 Score: 83.12%
• Initial target: 75% Weighted F1
• 12 classification categories
• TextCNN model built with PyTorch
• Full pipeline: data preparation, training, evaluation, deployment
• Interactive demo deployed on Hugging Face: claims-classifier-demo

This project demonstrates a concrete business use case: automatically classifying incoming customer requests in order to route them to the right category, team, or service.

In an operational context, this type of model could support:

• customer service teams
• logistics claims management
• transport dispute classification
• after-sales request routing
• ticket prioritization
• repetitive request processing

A Python command-line tool designed around synthetic supply-chain data.

Project README: TROEL OPS Kit

The project includes:

• data ingestion
• data validation
• KPI computation
• operational alerts
• reporting workflows
• simple and auditable logic

This project reflects my preferred approach: practical, transparent, and focused on operational usefulness.

Supply Chain · Logistics · Inventory Management · Warehouse Operations · Flow Management · Operational Coordination · KPI Monitoring · Stock Reliability · Process Improvement · SAP · WMS · ERP

Python · SQL · Data Analysis · Data Processing · KPI Reporting · Workflow Automation · Data Validation · Operational Monitoring

NLP · Machine Learning · Deep Learning · PyTorch · Text Classification · Model Evaluation · Hugging Face Deployment · Local RAG · AI Assistants · Decision Support

My working style is based on:

• field experience before theory
• business-first thinking
• simple and maintainable solutions
• clear operational ownership
• measurable performance indicators
• explainable logic
• practical deployment rather than demo-only projects

I believe that useful AI in companies should be close to the field, understandable by business teams, and connected to concrete operational problems.

Each project lives in its own folder.

GitHub Actions workflows are centralized at the repository root in:

.github/workflows/

• LinkedIn: christophe-troel
• Hugging Face: FrenchEdtech