
Table of Contents
Last update: 2023. All opinions are my own.
1. Overview
Collaborated with a team to predict raw material demand and improve inventory turnover for Industrias Duero.
2. Data Pipeline
We integrated supply-chain data sources and built a clean forecasting dataset, then validated assumptions with stakeholders.
3. Modeling
- Implemented time-series and ML forecasting strategies.
- Compared classical models with ML baselines.
- Designed evaluation to reflect operational lead times.
4. Outcomes
- Analyzed over 1.5 million data points across the supply chain.
- Built dashboards to communicate forecasting outcomes.
- Delivered recommendations for inventory turnover improvements.
5. Skills and Tools
- Python
- Facebook Prophet
- Time-series forecasting
- XGBoost, CatBoost
- Microsoft Power BI
