Automated Model Training Workflows with AWS S3

Motivation

Consider you’re an e-commerce platform aiming to enhance recommendation personalization. Your data resides in S3.

To refine recommendations, you plan to retrain recommendation models using fresh customer interaction data whenever a new file is added to S3. But how exactly do you approach this task?

This repo shows how you can set up an automated model training workflows triggered by S3 using Kestra.

Set up

1. Clone this repo:

git clone https://github.com/kaushikrohit004/MLOps-Workflow.git

2. Go to the folder:

cd MLOps-Workflow

3. Create the ".env" file and pass your GitHub and AWS credentials:

GITHUB_USERNAME=mygithubusername
GITHUB_PASSWORD=mygithubtoken
AWS_ACCESS_KEY_ID=myawsaccesskey
AWS_SECRET_ACCESS_KEY=myawssecretaccesskey
# ! This line should be empty

Next, encode these secrets using the following bash script:

bash encode_env.sh

Executing this script generates a “.env_encoded” file containing encoded secrets:

# .env_encoded
SECRET_GITHUB_USERNAME=bXlnaXRodWJ1c2VybmFtZQ==
SECRET_GITHUB_PASSWORD=bXlnaXRodWJ0b2tlbg==
SECRET_AWS_ACCESS_KEY_ID=bXlhd3NhY2Nlc3NrZXk=
SECRET_AWS_SECRET_ACCESS_KEY=bXlhd3NzZWNyZXRhY2Nlc3NrZXk=

4. Start the Kestra server using Docker Compose

docker-compose up -d