Building a solid foundation for MLOps: data labeling, versioning and validation

Hey all, we're getting back to our monthly event I'm happy to announce our next meetup!

Abstract:
All AI projects start with data — no matter how simple your idea is, you cannot develop machine learning algorithms without examples to train them on. But how do you create high-quality datasets? In this talk, we will walk you through all the steps.

Data labeling - Magdalena Konkiewicz, Data Evangelist, Toloka
The process usually starts with data annotation and requires building reproducible data labeling pipelines. This is a large-scale people management task where quality control becomes a crucial component of the project. Toloka is a crowdsourcing platform that can be used to build scalable and effective data labeling pipelines and we will show you how to take advantage of this tool.

Data Management - Moses Guttmann, CEO, ClearML
While high quality labels are a must, data is an ever-changing entity and proper data management is crucial to keep track of changes. We’ll discuss common challenges and how to overcome them using ClearML-Data.

Data Validation - Noam Bressler, ML Team Lead, Deepchecks
Every piece of traditional software goes through comprehensive tests of various types before deployment, and is continuously monitored in production. How can we adapt these ideas to the data-oriented world of ML? We’ll discuss best practices for extensively testing and analyzing data and models using deepchecks. Using the open source package, we'll show how to detect errors in your data, model, and methodology both before an ML product is released and throughout its production lifecycle.

The meetup is a hybrid meetup, come join us at Armis office, 8 Izchak Sade, Tel Aviv-Yafo
Or see the live stream @ https://toloka-ai.zoom.us/j/84569562548?pwd=aW54d29oUGFUa1daTko3N2hPQmdQdz09

Schedule:
18:00 - Meet, Greet, Eat pizza and drink beer
18:30 - Data Labeling - Magdalena Konkiewicz.
19:00 - Data Management - Moses Guttmann
19:30 - Data Validation - Noam Bressler