Take your machine learning model out of your desk drawer and show its benefit to the world through a simple API using AWS Lambda and API gateway. This tutorial will bridge the gap between having a machine learning model (e.g. in your Jupyter notebooks) and taking it to a level where others can benefit from it (i.e. through an API).
Agenda: - Why you need a service for your model? - Short intro to AWS Lambda and API gateway - Constructing a service around your model - Setting up AWS & deployment environment - Using the serverless framework to deploy your service to AWS - Things you should be aware of
Prerequisites:
You should be familiar with using Docker and have it installed.
Please refer to the GitHub repo accompanying this workshop and
install the necessary software: bweigel/ml_at_awslambda_pydatabln2018.
Also, you should already have an Amazon Web Services (AWS) account if you want to follow this session.
We will be using the free tier of AWS.
Slides for the talk: https://bweigel.github.io/pydata_bln_2018/