Example how to use lambda and custom containers to create serverless microservice for ML Inference for CPU bound requests