Building an Open Source MLOps Stack with ZenML - Part 2

Explore the latest developments in ZenML, an open-source MLOps framework, in this 59-minute conference talk by co-founder Hamza Tahir. Dive into the complete re-architecture of ZenML, designed to streamline the transition from local development to production pipelines. Learn about the new features, including a brand-new UI, shared stacks, and centralized configuration. Watch a comprehensive demo showcasing ZenML's capabilities in local development, staging pipelines, and integration with GitHub actions. Gain insights into the framework's ability to create portable ML pipelines and its potential to standardize ML production processes. Discover how ZenML aims to solve common challenges in bringing machine learning into production and enhance collaboration among data scientists and ML engineers.

Syllabus

[] Introduction to Hamza Tahir
[] Big things to be shared on this episode
[] ZenML re-architecture / Introduction
[] ZenML early days
[] Bringing ML into production is still a broken process
[] Standardizing the process eases the pain
[] ZenML: Author portable ML pipelines
[] ZenML 0.20.0 Release
[] ZenML Architectural Change
[] Before 0.20.0 Remote case
[] After 0.20.0 Remote case
[] ZenML Dashboard: a beautiful, new look
[] Easily launch the dashboard from the CLI
[] The ZenML Dashboard is an Open-source
[] Easily deploy with ZenML CLI
[] Collaboration like never before
[] Shared stacks
[] Centralizing configuration
[] Demo
[] ZenML Pipelines Connect and Travel Across Stacks
[] Local Development with ZenML
[] Staging Pipeline
[] Coding
[] Integration between ZenML and Github actions
[] Connect with ZenML
[] ZenML Restarting process
[] Hamza's involvement in design decisions
[] Wrap up