001

/

Demand projection

/

Retail

Inventory planning

Demand-based material requirements planning system that minimizes the percentage of out-of-stock products and lost sales for the leading retailer in Colombia.

Databricks
Databricks
Databricks
Spark
Spark
Spark
Scikit-learn
Scikit-learn
Scikit-learn
Machine learning
Machine learning
Machine learning
checking-inventory
checking-inventory
checking-inventory

Challenge

An important retail company in Colombia, with more than 2,500 stores, faced the challenge of reducing key indicators such as out-of-stocks and lost sales to improve service and maximize sales. They sought to develop tools that would facilitate inventory management and strengthen the replenishment process through data-driven decisions.

Solution

At Sumz, we develop an advanced DDMRP system (Demand-Driven Material Requirements Planning) that integrates artificial intelligence to optimize inventory management. This innovative system predicts demand and generates daily order suggestions, adjusting to each specific store and material. Additionally, it considers process uncertainty and various factors influencing needs, such as demand seasonality, order frequencies, and space limitations.

We also accompanied the implementation of the system during a 90-day pilot, where we randomly selected a group of stores to implement and measure the results of the DDMRP system. To do this, we designed a tracking dashboard, where the impact of DDMRP on indicators such as out-of-stocks and lost sales was evidenced.

The representation of our data universe was created with the help of artificial intelligence and teamwork.

Carrera 13 # 93-68 Bogotá, Colombia

©2024

The representation of our data universe was created with the help of artificial intelligence and teamwork.

Carrera 13 # 93-68 Bogotá, Colombia

©2024

The representation of our data universe was created with the help of artificial intelligence and teamwork.

Carrera 13 # 93-68 Bogotá, Colombia

©2024