Kubernetes to Serve a Fraud Detection Model

Explore how Kubernetes can be leveraged to serve a real-time fraud detection machine learning application in this insightful conference talk. Discover the challenges and solutions in implementing a critical API for transaction processing, focusing on reliability, scalability, and performance. Learn about utilizing Kubernetes' scaling capabilities to handle transaction peaks without compromising model performance or response quality. Gain insights into setting up auto-deployments with automated rollbacks based on CI/CD end-to-end test results. Understand how to incorporate business logic using Kubernetes' Python client and pod management configurations. Delve into topics such as credit card fraud mechanics, chargebacks, machine learning features and training, order evaluation API development, autoscaling, and automated test suites. Enhance your knowledge of applying Kubernetes in a high-stakes financial technology environment.

Introduction
How credit card fraud works
Chargebacks
Machine Learning
ML Features
ML Training
Order Evaluation API
Building the API
Autoscaling
Automated test suites