Discover a framework for tackling MLOps challenges in this 40-minute conference talk from MLOps World: Machine Learning in Production. Learn how to navigate the complex landscape of MLOps tools and overcome "MLOps Fatigue" with Dean Pleban, Co-Founder & CEO of DagsHub. Explore first principles thinking, define problem parameters, and evaluate solutions through stress testing. Gain insights into data versioning and a 5-step process for selecting the best tools for your team's needs. Apply multi-disciplinary approaches to build effective products for data scientists and machine learning engineers.
Solving MLOps From First Principles - A Framework to Reduce Complexity
MLOps World: Machine Learning in Production via YouTube
Overview
Syllabus
Intro
First principles thinking.
4 Stages of Grief
Assumptions about the PROBLEM
Assumptions about the SOLUTION
Define the problem parameters
Google the problem
Step 4: Evaluate solutions - Stress Test
Example: Data Versioning
The 5 step process
Taught by
MLOps World: Machine Learning in Production
