Evolving Perspectives on Operational AI: MLOps with Full Stack Optimizations

Operational AI describes using models and algorithms to integrate AI into day-to-day customer experiences and business processes.

This session explores the concept generally, plus dives into a methodology for achieving it—combining MLOps^† components and AI optimizations to deploy performant and scalable AI solutions in production while ensuring that individual components of an AI system are optimized across the stack.

Key learnings:

• Practical examples of implementing MLOps components—such as model registries, data versioning, and monitoring—with AI tools such as model compression and Intel®-optimized AI frameworks
• How implementation of software and hardware optimizations can help optimize critical parts of the machine learning lifecycle
• How to use the combo of MLOps and AI optimization to maximize return on investment (ROI) and AI system quality

This session takes advantage of the latest Intel® hardware and software available in the Intel® Tiber™ AI Cloud. 

Skill level: Intermediate

Get the following tools as a stand-alone component or as part of the AI Frameworks and Tools.

Intel® Neural Compressor

Intel® Extension for PyTorch*

Intel® Extension for Scikit-learn*

Fine-Tuning Text Classification Model with Intel Neural Compressor

Get Started with Intel Extension for PyTorch

See All Code Samples

^†MLOps (machine learning operations) productionizes the operational rigor of the machine learning lifecycle.