From the course: Cloud Architecture: Design Decisions

Unlock this course with a free trial

Join today to access over 24,700 courses taught by industry experts.

Map requirements to an AI solution

Map requirements to an AI solution

From the course: Cloud Architecture: Design Decisions

Start my 1-month free trial Buy for my team

Map requirements to an AI solution

“

- Mapping requirements for a cloud-based AI solution follows a similar process to that outlined in this course for storage, compute, and databases. First, business requirements for the AI solution must be defined. Then, translate these into technical specifications considering the following, compute, storage, database, CloudOps, security, and governance. Let's go back and unpack these one at a time. Compute, determine the necessary CPU, memory, and operating systems such as GPU-optimized instances or deep learning and generative AI. Specify disaster recovery strategy, such as active-active and active-passive as covered in a previous video. Also provide detailed application interfaces, or APIs, that are used to interact with the AI models. Plan for the systems maintenance and configuration management. Implement robust security, including IAM and encryption. Storage, assess the capacity, meaning gigabytes or terabytes, needed for training data, models, and results. Choose appropriate…

Contents