Formation en IA & Data: Advanced MLOps: pipelines, monitoring, and scaling - Ascent Formation
Back to trainings
IA & Data

Advanced MLOps: pipelines, monitoring, and scaling

3 jour(s)21h

Description

Designing industrialized ML pipelines, effectively monitoring models in production, and implementing deployment and scaling strategies tailored to the performance, reliability, and cost constraints of an enterprise environment.

Learning Objectives

  • Design an end-to-end automated MLOps pipeline.
  • Orchestrate ML workflows with Kubeflow, Vertex AI Pipelines, and SageMaker Pipelines.
  • Implement a feature store and manage features in production.
  • Manage versioning, model registration, and promotion.
  • Detect data drift and model degradation.
  • Deploy A/B testing and progressive rollout strategies.
  • Optimize GPU allocation and scheduling on Kubernetes.
  • Select a serving strategy: batch or real-time.

Target Audience

ML Engineers
MLOps Engineers
DevOps Engineers
Data / AI / Cloud Architects
ML Platform Managers

Prerequisites

Strong proficiency in Machine Learning
Experience in Python and model deployment
Knowledge of Docker and Kubernetes
Familiarity with cloud environments or data pipelines

Program Outline

Informations

Duration

3 jour(s)

21h

Tarif

Sur demande

    Advanced MLOps: pipelines, monitoring, and scaling | Ascent Formation | Ascent Formation