This is the accompanying repo for a hands on workshop for building a ML infrastructure based on a real life system used in production.
The architecture is based on the PayPal fraud detection infrastructure and is designed for the conference attendees to have a hands on experience of a ML infrastructure.
The workshop was given at the Go Get conference in 05/2020 and will be given again at GopherCon in 11/2020.
The workshop is created by Natalie Pistunovich.
- Have GitHub set up on your machine
- Clone this repo
- Install Go (v1.14)
- Install the latest Docker
- Pull the Docker images:
- docker pull aerospike/aerospike-server:5.0.0.10
- docker pull aerospike/aerospike-tools:3.28.0
- docker pull tensorflow/serving:2.2.0
- Download the Kaggle dataset of credit card transactions
In case you are new to Go, here are some useful links:
- A summarized overview of Go: Learn Go in Y Minutes
- A short hands on tutorial: Tour of Go
- A handy cheatsheet: Go by Example