Server-side ML. Training & hosting for inference, with a goal towards serverless. AWS SageMaker, Batch, Lambda, EFS, Cortex.dev