Explore the challenges and solutions for deploying AI/ML applications in this 58-minute webinar on MLOps automation using Git-based CI/CD. Learn about ML pipeline workflows, collaboration between multidisciplinary teams, and automating the deployment process using cloud-native paradigms, Git, and Kubernetes. Discover how to maximize efficiency, leverage Git review processes for model evaluation, and simplify Kubernetes and DevOps complexities. Watch a demonstration of continuous delivery for machine learning in production environments using Git, CI frameworks, hosted Kubernetes, Kubeflow, MLOps orchestration tools, and serverless functions. Gain insights into real-world applications, including predictive maintenance and fraud detection, while understanding the importance of automation in bringing AI projects to production successfully.
Overview
Syllabus
Intro
80% of AI Projects Never Make it to Produc
Did you Try Running Notebooks in Product
Model and Code Development are Just the First Step
Example: Predictive Maintenance Pipeline
You can use Separate Tools & Services, Or you can use Kubernetes as the Baseline
What is an Automated ML Pipeline ?
Under The Hood: Open, Scalable, Production Ready
Serverless Simplicity, Maximum Performance
Serverless: Resource Elasticity, Automated Deployment and Operations
Dynamic Scaling for Intensive Workloads
KubeFlow: Automated ML Pipelines & Tracking
Simple, Production-Ready Development Process
Building CI/CD Process for ML(Ops)
Traditional Fraud-Detection Architecture (Hadoop)
Real-Time Fraud Prediction & Prevention
Taught by
CNCF [Cloud Native Computing Foundation]
