Explore a comprehensive podcast episode featuring Niels Bantilan, Chief Machine Learning Engineer at Union, discussing the crucial role of infrastructure in machine learning and leveraging open-source tools. Delve into topics such as scalable data processing, experiment tracking, hardware requirements, versioning tools, and monitoring systems for ML projects. Learn about orchestration tools like Flyte and how they unify various components of the ML lifecycle. Gain insights on MLOps challenges, the importance of reproducibility, and the evolution of ML infrastructure. Discover how mature companies are embracing Kubernetes and the future horizons of ML systems. Perfect for data scientists, ML engineers, and anyone interested in the intersection of machine learning and infrastructure.
The Role of Infrastructure in ML Leveraging Open Source - MLOps Podcast #197
Overview
Syllabus
[] Niels' preferred coffee
[] Takeaways
[] Shout out to our Premium Brand Partner, Union!
[] Pandera
[] Creating a company
[] Injecting ML for Data
[] ML for Infrastructure Optimization
[] AI Implementation Challenges
[] Generative DevOps movement
[] Pushing Limits: Code Responsibility
[] Orchestration in OpenAI's Dev Day
[] MLOps Stack: Layers & Challenges
[] Mature Companies Embrace Kubernetes
[] Horizon Challenges
[] Flexible Integration for Resources
[] MLOps Reproducibility Challenges
[] MLOps Maturity Spectrum
[] First-Class Citizens in Design
[] Delegating for Efficient Collaboration
[] Wrap up
Taught by
MLOps.community
Related Courses
-
MLOps vs ML Orchestration - Exploring Challenges and Community-Driven Development
-
MLOps at the Crossroads - Challenges and Future Directions
-
Fine-Tuning Language Models with Declarative ML Orchestration
-
Designing ML Infrastructure for ML and LLM Use Cases
-
Systems Engineering in Machine Learning - Navigating Low-Level Challenges
-
Unpacking Three Types of Feature Stores - MLOps Podcast
