Главная
Study mode:
on
1
Intro
2
ML driven experiences
3
"Hidden Debt" of ML
4
How can ML Practitioners focus on their craft?
5
Principles of Intuit ML Platform
6
Generic Model Lifecycle
7
Technologies in use
8
Feature Store
9
Feature Processing
10
Model Training
11
Automate the simple things
12
With great speed, comes cost
13
Cost transparency
14
What can go wrong?
15
Security
16
Compliance
17
Learnings
Description:
Explore Intuit's approach to managing machine learning models at scale in this 26-minute conference talk from OpML '20. Dive into the challenges of handling large, sensitive datasets and the continuous need for model training and tuning while maintaining high security and compliance standards. Learn about Intuit's Machine Learning Platform, which leverages GitOps, SageMaker, Kubernetes, and Argo Workflows to provide scalable and secure Model LifeCycle management capabilities. Discover how the platform addresses data science and MLE needs while meeting enterprise requirements, and examine key features such as feature management, processing, bill backs, collaborations, and separation of operational concerns. Gain insights into how these innovations have led to a 200% increase in model publishing velocity, and explore the technologies, principles, and learnings behind Intuit's successful ML model management strategy.

Managing ML Models at Scale - Intuit’s ML Platform

USENIX
Add to list