Major Soccer League Club

Introduction

A professional soccer club in Major League Soccer (MLS), is setting new standards in the league by incorporating data analytics into their decision-making processes. With Kestra as their workflow orchestration tool, the club is making great strides in scouting, player performance analytics, and financial strategy.

Technology Stack

This club use a comprehensive technology stack to execute its data-driven objectives:

• AWS S3: Cloud-based storage for raw and processed data.
• PostgreSQL: The club's primary database server.
• Shiny Proxy: Used for serving apps.
• Python and R Codebases: Employed for creating custom scripts and machine learning models.
• Kestra: Serves as the workflow orchestrator, overseeing the management of pipelines.

The Challenge

Before using Kestra, the club struggled with a labor-intensive data stack based on Papermill notebooks. Troubleshooting was tedious, and managing the complex code became a bottleneck for effective analysis. The club needed a more efficient and scalable solution to keep up with the rapid pace of MLS and its weekly sporting performances.

Why Kestra?

This MLS club chose Kestra for its:

• Comprehensive Control Plane: Provides a web browser-based interface to create flows, monitor executions, and build dependencies.
• Versatility: Can easily be used across different teams and is templatable, making it highly reusable.
• Polyglot Support: Capable of orchestrating workflows involving both Python and R codebases, thanks to its Docker container support.
• Ease of Integration: Simplifies the fetching of data from multiple sources and integrates effortlessly with AWS and PostgreSQL.

The Outcome

With the adoption of Kestra, they were able to:

• Streamline Data Ingestion: The club can now easily fetch data from multiple providers and orchestrate its ingestion into AWS S3 and PostgreSQL.
• Improve KPI Management: Kestra allows the club to create bespoke KPIs for match analysis and player recruitment.
• Enhance Scouting Processes: Incorporation of machine learning models into the scouting process has become straightforward.
• Increase Efficiency: The club's data analytics team can now focus on improving models and analytics instead of dealing with complex code and manual workflows.

Next Steps

Having successfully migrated to Kestra, the analytics team is now working on a possession value model that aims to quantify a player's contribution to the team in terms of goals added. They are keen on integrating data from their provider, StatsBomb, to improve the accuracy of their in-house models.