MLOps Foundations: 5-Week Implementation

MLOps Foundations on Azure is delivered through a custom engagement where we first determine your infrastructure needs and then tailor your solution to meet your model development, operations and compliance requirements.

Rackspace Model Factory Framework features:​

• Experiment tracking​
• Project packaging​
• Deploy models from anywhere​
• Tracking UI​
• Model storage​
• Support major ML frameworks​ on Azure
• CI/CD integration​

Engagement Activities

• WEEK 1: Discovery & Planning
  • Review current state and design environment deployment
  • Outcome – Environment deployment scope and design complete
• WEEK 2-3: Environment Implementation
  • Design, build, test, and validate environments and integrations with Azure, build ML pipelines and CI/CD integrations
  • Outcome – Dev, QA, and production environments complete
• WEEK 4: Integration & Onboarding
  • Onboard, integrate, test and validate models into the Azure ML pipelines
  • Outcome – 1-2 working models and pipelines
• WEEK 5: Knowledge Transfer
  • Provide Azure documentation and training to customer's teams
  • Outcome – Customer's data science teams enabled to onboard new models

Key benefits and business outcomes

• Automated model deployment across development, Q/A and production environments
• Reproducibility of models and predictions
• Platform for collaboration
• Standardized model development environment for your data science teams
• Model explainability for governance and regulatory compliance
• Tools for diagnostics, performance monitoring and addressing model drift
• Accelerated ROI