Building data science teams to have an impact at scale

This post details the integration of data science teams with infrastructure engineering teams at Meta to tackle large-scale challenges. It highlights specific contributions in: 1. CDN cache efficiency through ML models predicting future content requests. 2. Performance regression detection tools for both client-side and server-side applications, using custom algorithms to account for trends, seasonality, and noise. 3. Data center demand forecasting using custom Bayesian time series and deep learning models, and predicting the impact of infrequent events on infrastructure.