Explore the evolving landscape of MLOps, enterprise adoption, and the changing LLM ecosystem in this insightful podcast featuring Chris Van Pelt, CISO and co-founder of Weights & Biases. Gain valuable insights into effective ML evaluation processes, demo-driven development, and the complexities of MLOps pipelines. Discover Van Pelt's journey from Crowdflower to Weights and Biases, learning about the importance of early security integration in ML projects. Delve into discussions on the transformative impact of machine learning on the tech industry, challenges in detecting subtle bugs, and the potential of open-source models and multimodal capabilities. Benefit from expert advice on easy wins in MLOps implementation and understand the iterative value evaluation process for new releases.
Enterprises Using MLOps - The Changing LLM Landscape and MLOps Pipelines
Overview
Syllabus
[] Chris' preferred coffee
[] Takeaways
[] Huge shout out to Weights & Biases for sponsoring this episode!
[] Please like, share, and subscribe to our MLOps channels!
[] CrowdFlower
[] Difference of CrowdFlower and Trajectory
[] Transition from CrowdFlower to Weights & Biases
[] Excel spreadsheets being passed around via email
[] Evolution of Weights & Biases
[] CISO role
[] Advise for easy wins
[] Transition into LLMs
[] Prompt injection risks on data
[] LLMs for New Personas
[] Iterative Value Evaluation Process
[] Iterating on New Release
[] Evaluation survey
[] Landscape of LLMs and its evolution
[] Conan O'Brien
[] Wrap up
Taught by
MLOps.community
Related Courses
-
How to Systematically Test and Evaluate LLM Apps - MLOps Podcast
-
Evaluating LLMs with Weights & Biases and Docker Desktop
-
How Replit Trained Their Own LLMs - LLM Bootcamp
-
Aligning LLM Judges for Better Evaluations in RAG Pipelines
-
MLOps vs ML Orchestration - Exploring Challenges and Community-Driven Development
-
LLM Evaluation: Challenges and Best Practices - MLOps Podcast #210
