Home > Servers > Modular Servers > White Papers > Reference Architecture: Machine Learning Containers on PowerEdge MX and VMware Cloud Foundation 4.0 with Tanzu > USE CASE A: Marketing recommendation system
Container Stack: XGBoost, Kubeflow Python SDK, Kubeflow pipeline
Vertical: Retail, marketing
This is a train-and-serve model that allows you to create a simple prediction for marketing teams in a few code lines. It addresses the problem of finding optimal pricing for goods and products.
Pricing decisions are critically important for any business, as pricing is directly linked to demand and profits. Even one suboptimal decision can lead to tangible losses, and major mistakes can have grave consequences.
Although the fundamentals of price optimization are well understood, experience with leading retailers indicates that retail practitioners still struggle with pricing decisions. Even those who presumably make the right pricing decisions are often concerned they have unharvested profits or avoidable losses due to suboptimal pricing, and the lack of tools and techniques to measure it quantitatively.