Hamza Tahir and Adam Probst on the state of MLOps and tools for productionizing machine learning pipelines.
Subscribe: Apple • Android • Spotify • Stitcher • Google • AntennaPod • RSS.
Hamza Tahir and Adam Probst are co-creators of ZenML, an extensible open source framework for building reproducible pipelines. We discuss the current state of ZenML, the many use cases that ZenML has been designed for, and its near-term roadmap. We also dive into MLOps (trends, challenges, and opportunities) as well as the ecosystem of tools and processes for productionizing machine learning pipelines.
To learn more about Ray and how to scale machine learning applications, attend Ray Summit (San Francisco / Aug 23-24)
Highlights in the video version:
- Origin story of ZenML and target persona
ZenML – a framework that integrates and optimizes pipelines
Importance of an orchestrator
Solving data pipeline issues
Challenges of an exponential number of tools
What’s on the roadmap or next priority?
Data annotation
Integrating with non open source tools
Integration testing from a software engineering perspective
Upcoming ZenML events
Related content:
- A video version of this conversation is available on our YouTube channel.
- Experiment Tracking and Experiment Management Tools
- Machine Learning Trends You Need To Know
- Here’s what you need to look for in a model server to build ML-powered services
- Jules Damji and Richard Liaw: Scalable, end-to-end machine learning, for everyone
- Nikhil Muralidhar: MLOps Anti-Patterns
- Hamel Husain: Deploying Machine Learning Models Safely and Systematically
- Jeremiah Lowin: Dataflow Automation
- Nick Schrock: Software-Defined Assets
- Haytham Abuelfutuh: Orchestrating Machine Learning Applications
Attend Ray Summit: Save 25% on your pass with the discount code Ben25
[Image: Window Front at Night from Wikimedia.]
