MLOps with Red Hat OpenShift AI following GitOps

This repository is a comprehensive framework for deploying and managing machine learning projects on Red Hat OpenShift AI using a GitOps approach. It provides tools and examples to streamline the deployment of AI/ML workloads, including a standard Helm chart for data science projects and a reference example for deploying a fraud detection model.

Features

1. Red Hat OpenShift AI Installation (GitOps Mode)

• Leverages GitOps principles to automate the installation and management of AI tools on OpenShift.
• Ensures consistency and traceability across environments.

2. Standard Helm Chart for Data Science Projects

• A reusable and configurable Helm chart designed for deploying various data science workloads.
• Simplifies deployment and scaling of ML models and supporting infrastructure.

3. Fraud Detection Model Example

• Demonstrates a complete workflow for deploying a fraud detection model.
• Implements GitOps practices to manage the lifecycle of the model.
• Includes best practices for CI/CD pipelines.

Getting Started

Prerequisites

1. Red Hat OpenShift
   • Assume you already have OpenShift 4.17 or later installed.
   • Install the OpenShift CLI (oc).

Installation

Step 1: Clone the Repository

git clone https://github.com/davidseve/mlops.git
cd mlops

Step 2: Install OpenShift AI

cd bootstrap
./bootstrap.sh

It could take several minutes

Step 3: Validate dsc

oc get deployments -n redhat-ods-applications

Step 4: Create AI Fraud Detection Example

cd ../ai-examples/fraud-detection/testing
./test-fraud.sh

It could take several minutes

Step 4: Validate Fraud Detection inference

host=<YOUR_HOST>
url="https://fraudinference-fraud.apps.$host/v2/models/fraudinference/versions/1/infer" 
data='{
        "id" : "42",
        "inputs": [
                    {
                        "name": "dense_input",
                        "shape": [1, 5],
                        "datatype": "FP32",
                        "data": [0.3111400080477545, 1.9459399775518593, 1.0, 0.0, 0.0]
                    }
                ]
        }'
curl -k -X POST "$url" -H "Content-Type: application/json" -d "$data"

Clean up

Step 1: Delete AI Fraud Detection Example

./delete-fraud.sh

Step 2: Delete Install OpenShift AI

cd ../../../bootstrap/
./delete.sh