MLOps & Data Synthesis

DATA SCIENCE CLUB (powered by МАШИНЕРИЈА) announces its third meetup in Novi Sad! This time we prepared two very interesting talks. Our speakers will be Mikhail Rozhkov and Roman Steinberg.

When: Saturday, March 18, 16:00 - 19:00 CET
Where: Startit center, Novi Sad, Miroslava Antića 2

First talk:
Build Data validation and Model Monitoring pipelines with DVC and Evidently

Abstract
Data validation and model monitoring are critical aspects of any machine learning project. Data validation involves verifying that the data used for training and testing a model is accurate, complete, and consistent with the intended use case. Model monitoring involves tracking the performance of a model over time to ensure that it continues to meet the desired performance criteria and does not drift from its original performance.
In this talk we will explain the importance of data validation and model monitoring in machine learning projects and demonstrate how to use DVC and Evidently to implement robust data validation and model monitoring pipelines.

Speaker:
Mikhail Rozhkov
ML engineer and MLOps Solutions Architect
Mikhail Rozhkov is an ML engineer and MLOps Solutions Architect with over 7 years of experience in Machine Learning and Data Science. He is a founder of the ML REPA community (mlrepa.org), and author of courses on ML automation, DVC, and MLOps topics.
Helped 200+ engineers learn advanced concepts and tools for ML engineering and MLOps. Collaborated with 50+ companies and ML teams in US, Europe, and Asia to design MLOps processes and integrated open-source tools.
LinkedIn:[ https://www.linkedin.com/in/mnrozhkov](https://www.linkedin.com/in/mnrozhkov)
GitHub:[ https://github.com/mnrozhkov](https://github.com/mnrozhkov)

Second talk:
Can't collect data - generate it

Abstract
This talk will focus on the lack of data problem and one of the ways to solve it - data synthesis. In the beginning, we will discuss some cases and interesting papers which use data synthesis. Finally, we consider an example of a real project in one of the banks, where it was necessary to train neural networks on synthetic data.

Speaker:
Roman Steinberg PhD, CTO & co-founder at Vision Systems with 25 years in IT Industry. Last 8 years in ML/DL.