Deploy a custom model with AWS SageMaker and build a Front-end with S3

This repo contains code for the tutorial article I made on Medium found here. The article shows how to deploy a custom model with Docker and SageMaker and create a serverless S3 front-end.

Repo contents:

• docker-image-code/: contains the code for the custom docker model image. Based on a template that works with the AWS SageMaker estimator wrapper
• front-end/: contains the code for the S3 website that allows users to interact with my google trends anomaly detection model
• deploy-custom-model.ipynb: The SageMaker notebook for deploying the model
• lambda-code.py: AWS Lambda function code for invoking the model endpoint

Go to my S3 website here to upload google trends CSV data and identity anomalies. (Note: it may not be up all the time)