As machine learning initiatives become more prominent across companies looking to leverage their data to improve future projections and decision-making, demand for frameworks that simplify ML model development has been soaring. In part 1 of this series, we looked at the challenges faced in ML model development and deployment that have resulted in the failure of more than 25% of AI & ML initiatives, as noted by IDC. We also discussed some options to improve the speed and ease of ML model development, from tools such as the Amazon SageMaker stack to the concept of enhancing operational efficiency across organizations.
In the second part of this series, we will take a look at Rackspace Technology’s Model Factory Framework(MLOps) and how it improves efficiency and speed across model development, deployment, monitoring, and governance.
End-to-End ML Blueprint
As we discussed earlier, a large variety of tools and frameworks exist within the Data Science/Machine Learning universe. When in development, ML models flow from data science teams to operational teams, and these preferential variances can introduce a large amount of lag in the absence of standardization.
The Rackspace Technology Model Factory Framework provides a model lifecycle management solution in the form of a modular architectural pattern built using open source tools that are platform, tooling, and framework agnostic. It is designed to improve the collaboration between data scientists and operations teams so that they can rapidly develop models, automate packaging, and deploy to multiple environments.
The framework allows integration with AWS services and industry-standard automation tools such as Jenkins, Airflow, and Kubeflow. It supports a variety of frameworks such as TensorFlow, scikit-learn, Spark ML, spaCy, PyTorch, etc., and can also be deployed into different hosting platforms such as Kubernetes or Amazon SageMaker.
Benefits of the Model Factory Framework
The Model Factory Framework affords large gains in efficiency, cutting the ML lifecycle from the average 15+ steps to as few as 5. Employing a single source of truth for management, it also automates the handoff process across teams, simplifies maintenance, and troubleshooting.
From the perspective of data scientists, the Model Factory Framework makes their code standardized and reproducible across environments, enables experiment and training tracking, and can result in up to 60% of compute cost savings as a result of scripted access to spot instance training. For operations teams, the framework offers built-in tools for diagnostics, performance monitoring and model drift mitigation. It also offers a model registry to track models’ versions over time. Overall, this helps the organization improve their model deployment time and reduce effort, accelerating time to business insights and ROI.
