The TWIML AI Podcast (formerly This Week in Machine Learning & Artificial Intelligence) Do You Dare Run Your ML Experiments in Production? with Ville Tuulos - #523
Sep 30, 2021
Ville Tuulos, CEO and co-founder of Outerbounds, shares his journey from building Metaflow at Netflix to writing his upcoming book on effective data science infrastructure. He discusses the unique challenges of ML production, emphasizing the need for innovative solutions and robust experimentation techniques. Ville highlights Metaflow's role in enhancing productivity and its integration with platforms like Kubernetes. The conversation also delves into the evolution of MLOps, especially in a post-pandemic world, and the growing community around these tools.
AI Snips
Chapters
Books
Transcript
Episode notes
Metaflow at Netflix
- Ville Tuulos led the machine learning infrastructure team at Netflix.
- They developed Metaflow to solve the challenge of deploying machine learning models in production.
Outer Bounds' Mission
- Outer Bounds is focused on commercializing Metaflow.
- They aim to help data scientists become more productive.
Metaflow's Purpose
- Metaflow addresses the difficulty of building production-ready ML applications.
- It simplifies the process by providing a user-friendly Python library.
