Sponsored Session: MLOps: Why DevOps Solutions Fall Short in the Machine Learning World – Eduardo Bonet, GitLab
While machine learning (ML) has been around for decades, we’ve only recently started looking at ML as software that we need to deploy, maintain, and iterate. MLOps is a young discipline that is still maturing. But why is it necessary in the first place? Why can’t we just use existing DevOps solutions? And what about large language models (LLMs) and LLMOps? In this session we will explore:
How classic software development is different from developing software containing ML
The implications of these differences, and what is needed to improve iteration time on MLOps
GitLab’s approach to bridge the gap between DevOps and MLOps
LLMOps, and why MLOps solutions fall short in the LLM world
