End-to-End-MLOps-Training-for-Insurance-Claims

Welcome to the End-to-End MLOps Training for Insurance Claims repository! This repository is designed to provide comprehensive training and resources for implementing End-to-End Machine Learning Operations (MLOps) workflow using Azure Machine Learning and Azure DevOps Platform.

• data/:

  • insurance.csv/: This CSV file contains the data for the MLOps project. This data was taken from the Kaggle and due to the cost efficiency, the study uses a subset of insurance data.

• deployment/:

  • aciDeploymentConfigStaging.yml/: The YAML file contains the scripts for configuring the Azure Container Instance used in the CD Staging Pipeline.
  • aksDeploymentConfigProd.yml/: The YAML file contains the scripts for configuring Azure Kubernetes Services used in the CD production Pipeline.
  • inferenceConfig.yml/: The YAML file contains the scripts for configuring Azure Kubernetes Services, scripts for running the model and the dependencies needed for the production environment.
  • score.py/: Python file contains scripts for loading the model and set-up for prediction.
  • scoringConfig.yml/: The YAML file contains the dependencies for the ACI image.

• environment_setup/:

  • cloud-environment.json: This JSON file is the Infrastructure as a Code for all the necessary resources for the MLOps Lifecycle.
  • iac-create-environment-pipeline-arm.yml: The YAML file contains all the resources and environment names as variable groups and values which will be used to connect and authenticate the resources using Azure Agents in the Continuous Integration Pipeline.

• package_requirement/:

  • install_requirements.sh: This bash script will read the requirements.txt and install all the packages.
  • requirements.txt: This text file contains all the versions of the packages.

• tests/integration:

  • conftest.py: Python scripts for setting up the pytest for testing the model.
  • prod_test.py: Python scripts for testing the model on the production environment.
  • staging_test.py: Python scripts for testing the model on the staging environment.

• training/:

  • conda_dependencies.yml: The YAML file specifying the Conda environment dependencies required for the project.
  • parameters.json: The JSON file containing parameters for the model.
  • train.py: Python script for train_test_split, train the model and evaluation setup for the machine learning model.
  • train_aml.py: Python script for training the Machine Learning model using Azure Machine Learning.
  • train_insurance.runconfig: This file is a configuration script for Azure Machine Learning experiments, defining compute resources, environment settings, and data references.
  • train_test.py: Python scripts for running unit tests in the DevOps Continuous Integration Pipeline.

• parameters.json/: The JSON file containing parameters for the model.