Explore a 50-minute podcast episode featuring Jeremy Thomas Jordan, a machine learning engineer at Duo Security, discussing the construction of threat detection systems from an MLE's perspective. Gain insights into the importance of writing as a learning tool, the use of rule engines in threat detection, managing rule sprawl, and effective testing for machine learning systems. Discover Jordan's approach to fundamentals, his blogging process, and the benefits of writing for personal development. Learn about the upcoming Normcore Conference and how to connect with the MLOps community through various platforms.
Building Threat Detection Systems - An MLE's Perspective
Overview
Syllabus
[] Jeremy's preferred coffee
[] Introduction of hosts
[] Introduction to Jeremy Thomas Jordan
[] Takeaways
[] Like, subscribe, and leave a review on our platforms
[] Normcore Conference
[] When are rule engines good and when are they not?
[] Going with Implemental Rule
[] Implementing a Rule
[] Managing rules sprawl
[] Reviewing and updating the rules
[] Chaining
[] Rules catching complicated threats
[] Notion of risk and confidence
[] Effective testing for machine learning systems blog post
[] Testing large language models
[] Model evaluation vs model testing
[] Approach to fundamentals
[] “The Cobbler’s Children Have No Shoes”
[] Writing being a checkpoint for Jeremy
[] Posting Jeremy's blogs on https://ghost.org/
[] Jeremy's blogs distribution platforms
[] Jeremy's most popular blogpost
[] Sign up for NormConf for free here https://normconf.com/ and attend on December 15th!
[] Wrap up
Taught by
MLOps.community
