Optimizing through Automation End to End Data Quality Checks and Model Deployment Feat Spectrum Labs скачать в хорошем качестве

Optimizing through Automation End to End Data Quality Checks and Model Deployment Feat Spectrum Labs

This talk features Nina Lopatina, Director of Data Science and Yuanbo Wang, Principal ML Engineer -- both of Spectrum Labs. About their talk: Often times, an ML-oriented organization reaches a tipping point when they can no longer overlook the importance of scaling up ML life cycle. It demands a common framework and platform to reduce time spent by data scientists on manual or repeated tasks. Although dozens of MLOps solutions exist, adopting them can be confusing and requires customization given the existing systems. In this talk, we will showcase an example of building out the first MLOps pipeline in a fast growing organization. We will discuss how we streamline the workflow, from quality checking the data to deploying and monitoring the models. We will also share trade offs around technical decisions to build pipelines that make the most sense for us. This talk was originally delivered at Arize:Observe 2023, a conference on the intersection of large language models, generative AI, and machine learning observability in the era of LLMops. Get updates from Arize on future events: https://arize.com/community/ Get certified in ML observability: https://courses.arize.com