Mark McQuade and Tanya Vucetic - Automating Production Level Machine Learning Operations on AWS
Speaker Bio -
Mark McQuade is a Practice Manager - Data Science and Engineering at Onica - a Rackspace Company.
Tanya Vucetic is a Data Scientist at Onica - a Rackspace Company.
Talk Abstract -
Machine Learning (ML) has revolutionized how we’ve solved business problems over the last decade. The ability to collect and store limitless data, coupled with advancements in computing and networking, has led to the use of Machine Learning in many business verticals.
However, developing end to end machine-learning pipelines and workflows that provide continuous and adaptive business insights to other applications or users is a challenge. This is primarily because of an inherent gap in how data scientists develop the machine learning models and how ML operations teams promote and deploy them into the production environments. Furthermore, complexities of CI/CD in the ML context, such as model governance and quality assessment, distinguish ML Ops from traditional DevOps. We will explore these specific challenges, and illustrate how familiar cloud services can be stitched together to bridge this gap between development and deployment, and to address the specific needs of ML Ops. The overall architecture pattern of a “model factory” enables support for numerous machine learning models in production and development simultaneously along with CI/CD for data science and automated workflows for Development, QA, and Production.
What we'll cover:
The gap between the Data Scientists and ML Operations
Why ML Ops is not DevOps
Architecture patterns necessary for elements of effective ML Ops
How a “model factory” architecture holistically addresses CI/CD for ML
Model Factory Demo that will explore:
Quick feedback and traceability for model development
ML framework-agnostic tooling for packaging of models
Platform agnostic continuous/rolling deployment
Watch video Mark McQuade and Tanya Vucetic - Automating Production Level Machine Learning Operations on AWS online, duration hours minute second in high quality that is uploaded to the channel Toronto Machine Learning Series (TMLS) 16 August 2023. Share the link to the video on social media so that your subscribers and friends will also watch this video. This video clip has been viewed 102 times and liked it 1 visitors.
