Scaling ML Serving to 1000s of Models ↗

My talk at DASH on scaling ML model serving infrastructure to handle thousands of models in production.