In this session, we are joined by Product Madness.com (an Aristorcat.com company):
Michael Leznik - Head of Data Science
Matthieu Glotz - Data Scientist
and Iguazio's (https://www.iguazio.com/)
Yaron Haviv - CTO & Co-Founder

We discuss how technology and new work processes can help the gaming and mobile app industries predict and mitigate 1st-day (or D0) user churn in real time — down to minutes and seconds using modern streaming data architectures such as KAPPA. Also, we explore feature engineering improvements to the RFM (Recency, Frequency, and Monetary) churn prediction framework: The Discrete Wavelet Transform (DWT).

00:00 - 02:53 - Intro
02:54 - 08:13 - MLOps overview
08:14 - 15:04 - Churn in Free2Play Games
15:05 - 24:02 - RFM + Discrete Wavelet Transform (DWT)
24:03 - 32:45 - Real-Time Feature Engineering
32:46 - 35:51 - Kappa Architecture
35:52 - 42:38 - Serverless Streaming
42:39 - 51:30 - DEMO
51:31 - 58:06 - Q&A
58:07 - 1:01:55 - Outro & Poll Question Results

The MLOps Live Webinar Series is a collection of bi-weekly online events during which data science leaders explore the elements of managing and automating machine learning pipelines to bring data science into real business applications. The sessions go beyond theory, with industry leaders sharing challenges and practical solutions.

Dont miss - MLOPs Live #6 - With Ecolab - Breaking the Silos Between Data Scientists, Eng & DevOps- https://www.youtube.com/watch