Course Modules

  • Define the motivations and goals of AI applications.
  • Identify opportunities for material value through the implementation of AI in a business process.
  • Define how AI contributes to the architecture of the application design.
  • Identify opportunities for optimization in AI systems.

  • Describe the types of design patterns used in AI applications.
  • Identify when to use a design pattern based on the application workload or goal.

  • Design and implement an API using popular Python frameworks. 
  • Use HTTP protocols, and build and use REST APIs.
  • Define interactions between servers and clients in modern software applications.

  • Build software application architecture diagrams (inclusive of an appropriate software design pattern).
  • Audit an architecture diagram to implement an AI solution based on best practices. Define which parts are responsible for client/server interactions, and recognize the other critical components of the application.
  • Design cloud-native AI applications based on Amazon Web Services (AWS)*.

  • Define the value and components of MLOps.
  • Implement data pipelines, model registries, observability and triggering, and version control.
  • Build MLOps components using popular open source frameworks (MLFlow and Kubeflow*).
  • Identify opportunities for MLOps components to improve performance, quality, and user experience in real-world applications.

  • Define the requirements of different AI workloads and how they impact application design.
  • Design applications for high-inference throughput.
  • Design applications for distributed training.
  • Activate hardware-level accelerations that include Intel-optimized software.
  • Use OpenMP, numactl, and Intel® oneAPI Math Kernel Library (oneMKL) to optimize the use of underlying hardware.
  • Perform basic performance profiling of AI applications.
  • Identify the correct opportunities for using a general-purpose compute system versus an accelerator.
  • Identify best practices for MLOps, hardware selection and workload management for the needs of a particular use case.

  •  Build a project during a live workshop where you practice and apply the concepts learned in previous modules.