Case Study: a professional sports analytics firm cuts machine learning pipeline runtime from 4.5 days to under 90 seconds with Six Feet Up

a professional sports analytics firm slashes ML pipeline runtime from 4.5 days to under 90 seconds with Six Feet Up

The customer, a professional sports analytics firm, faced the challenge of a machine learning pipeline that took four and a half days to process player performance data, which was far too slow for real-time analysis. They engaged vendor Six Feet Up to streamline the pipeline using AWS and Python, reduce its run time, and move the process to a serverless model to save time and money.

Six Feet Up analyzed the existing code and parallelized the pipeline's processes using AWS Lambda. They separated the code into individual elements, spun up parallel machines, and optimized large libraries. As a result, Six Feet Up reduced the pipeline's run time from over four days to under 90 seconds, providing the firm with a primary data source for validating game models faster than ever before.