MLOps (Machine Learning Operations): 4-Week Envisioning and Design Assessment

Spyglass’ MLOps (Machine Learning Operations) is a 4-Week envisioning and design assessment that is the first step on your journey to an enterprise-wide strategy for productionizing, monitoring, and managing your entire machine learning model lifecycle on Microsoft Azure.

This engagement will help you understand how to leverage MLOps in Azure to improve real business outcomes, by ensuring that models your data science teams build can truly serve your organization, instead of remaining as virtual laboratory experiments. We will also work with you, using a series of workshops and assessments, to design an MLOps strategy tailored to your specific organization and requirements.

Spyglass provides an integrated team of an ML architect, DevOps architect and platform architect to ensure that the envisioning results in a well-architected framework that follows Responsible AI principles on Azure Services.

AGENDA Weeks 1-2: Discovery, complete assessments, and workshops with your business, IT and data science teams Week 3: Develop architectures, strategy, and ML model governance as needed Week 4: Develop roadmap and migration plan, present to stakeholders

DELIVERABLES • Executive briefings, outlining the specific MLOps requirements and opportunities • MLOps envisioning architecture design documents • Deployment roadmap for the platform and new use cases on Azure • Migration plan to onboard existing ML workloads