For ML Engineers & Data Scientists

Documentation for ML Engineers & Data Scientists

Documentation for teams who build, deploy, and manage machine learning pipelines using the KFP Operator.

Who This Is For

This documentation is for you if you:

• Develop and train machine learning models
• Run experiments with different algorithms, hyperparameters, and datasets
• Deploy and automate ML workflows from data ingestion to model serving
• Manage model versioning, validation, and deployment

Prerequisites

Required Knowledge

• Basic machine learning concepts
• Python programming
• Basic Kubernetes (pods, services, kubectl)

Technical Requirements

• Kubernetes cluster with KFP Operator installed
• kubectl configured to access your cluster
• Container registry access (Docker Hub, GCR, etc.)
• ML framework (TFX, Kubeflow Pipelines, or similar)

Optional

• Docker for building custom pipeline images
• Git for version controlling pipeline code

Quick Start Checklist

• Verify access to your Kubernetes cluster
• Confirm KFP Operator is installed and running
• Set up your development environment
• Choose a tutorial based on your use case
• Deploy your first pipeline
• Explore additional examples and patterns

Common Use Cases

Model Training & Experimentation

• A/B testing different model architectures
• Continuous model retraining on new data
• Distributed training across multiple nodes

Model Deployment & Serving

• Integration with serving platforms

MLOps & Automation

• Event-driven pipeline execution
• Integration with CI/CD systems
• Multi-environment promotion workflows

Getting Help

• Troubleshooting - Common problems and solutions
• Best Practices - Recommended patterns
• GitHub Issues - Known issues
• GitHub Discussions - Community help
• Platform Engineer docs - Installation/configuration issues

Related Resources

• TFX Documentation - TensorFlow Extended framework
• Kubeflow Pipelines - ML workflow platform
• Kubernetes Documentation - Container orchestration

Next: Getting Started guide.