Objective
Upon completing this lab, you will be able to:
- Understand the core concept of Cloud FinOps (Financial Operations).
- Create a "back-of-the-envelope" cost estimate for a hypothetical AI application.
- Compare the quantitative (direct costs) and qualitative (maintenance, flexibility) trade-offs between FaaS (Function-as-a-Service) and IaaS (Infrastructure-as-a-Service) models for AI.
- Identify the scale threshold where one model becomes more cost-effective than the other.
- Analyze key strategic factors like switching costs and upgrade paths in infrastructure decisions.
Tools/Resources Needed
- A web browser with an internet connection.
- A spreadsheet tool (Google Sheets, Excel) or a document editor for calculations.
- Pricing Pages for Reference:
Description of Assignment
Background
Welcome to the world of Cloud FinOps, the practice of bringing financial accountability to the variable spending model of the cloud. As discussed in Chapter 7, choosing your cloud infrastructure is a critical strategic decision that involves trade-offs across the entire DIMASI Cube. This lab puts you in the role of a decision-maker and directly applies the "Build, Buy, or Borrow" framework.
- The FaaS scenario represents the "Buy" path—purchasing access to a finished capability through an API.
- The IaaS scenario represents the "Borrow" path—using an open-source model but taking full responsibility for managing the infrastructure to run it.
- The true "Build" path—training a foundational model from scratch—is financially and technically unrealistic for nearly any company that isn't a hyperscale tech giant. This lab focuses on the most common strategic decision today: whether to "Buy" a managed service or "Borrow" an open-source tool.
Your task is to compare these two paths, weighing the simplicity of the FaaS ("The Vending Machine") against the control of the IaaS ("The Digital Landlord"), and to explore how growth over time might shift the balance.