Get in Touch
 Duration 14 hours

Course Outline

Introduction to AI Inference using Docker

  • Comprehending AI inference workloads
  • Advantages of containerized inference
  • Deployment scenarios and associated constraints

Constructing AI Inference Containers

  • Choosing appropriate base images and frameworks
  • Packaging pre-trained models effectively
  • Organising inference code for efficient container execution

Securing Containerised AI Services

  • Reducing the container attack surface
  • Handling secrets and sensitive files securely
  • Strategies for safe networking and API exposure

Techniques for Portable Deployment

  • Optimising images to enhance portability
  • Guaranteeing predictable runtime environments
  • Managing dependencies consistently across platforms

Local Deployment and Testing

  • Executing services locally via Docker
  • Troubleshooting inference containers
  • Evaluating performance and reliability

Deployment on Servers and Cloud VMs

  • Adapting containers for remote environments
  • Setting up secure server access
  • Deploying inference APIs on cloud VMs

Utilising Docker Compose for Multi-Service AI Systems

  • Coordinating inference with supporting components
  • Managing environment variables and configurations
  • Scaling microservices using Compose

Monitoring and Maintaining AI Inference Services

  • Approaches to logging and observability
  • Identifying failures within inference pipelines
  • Updating and versioning models in production environments

Wrap-up and Future Directions

Requirements

  • A solid grasp of fundamental machine learning concepts
  • Practical experience with Python or backend development
  • Basic familiarity with core container concepts

Intended Audience

  • Developers
  • Backend engineers
  • Teams responsible for deploying AI services

Testimonials (2)

Upcoming Courses

Related Categories