MLOps.community  cover image

Towards Observability for ML Pipelines // Shreya Shankar // MLOps Coffee Sessions #75

MLOps.community

00:00

Ama Observability

A lot of people have come to me recently, for whatever reason, maybe because i was boutcast, and said, what is it with all monitoring stuff? There're so many companies out there. I know that i probably should do it, bake it into my aow m el system, but i'm not really sure how. And this is already a hard problem, right? Because we have things like label lag. We have a imperfect joins. It depends on how our monitoring is done. In streaming systems, we lose data of were windows. So the first question then is, what is the first thing that people monitor? Right? And that's where everyone is getting it wrong.

Transcript
Play full episode

The AI-powered Podcast Player

Save insights by tapping your headphones, chat with episodes, discover the best highlights - and more!
App store bannerPlay store banner
Get the app