Study AIF-C01 AI Use Cases and Technique Selection: key concepts, common traps, and exam decision cues.
AIF-C01 often asks a simpler question than people expect: is AI or ML appropriate here at all, and if so, what broad technique fits? Strong answers match the business need to the right category before they worry about AWS service names.
| Need | Strongest first fit |
|---|---|
| Predict a numeric value | regression |
| Sort into labeled categories | classification |
| Group similar items without labels | clustering |
| Generate text, images, or summaries | generative AI |
| Extract insight from large natural-language content | NLP-based or GenAI-assisted pattern |
One of the easiest ways to miss AIF-C01 is to force AI onto a problem that really wants deterministic rules, search, or a simple business workflow.
| If the problem is mainly about… | Better instinct |
|---|---|
| fixed business logic with clear rules | start with deterministic rules rather than ML |
| known labels and a prediction target | supervised ML is a stronger fit |
| no labels but similarity or segmentation goals | clustering may be the better lane |
| drafting or transforming natural-language output | GenAI becomes stronger |
| finding relevant information in trusted documents | search or retrieval may beat full generative behavior |
If the problem has historical examples with known answers, supervised ML often becomes stronger. If the problem is open-ended content generation, GenAI becomes stronger. If the task can be solved with rules more cheaply and reliably, AI may not be the best answer at all.
AWS wants you to classify the problem type before you classify the service:
A retailer wants to segment customers into similar groups for marketing, but it does not have predefined customer labels. Which technique is the strongest first fit?
Correct answer: C. Grouping similar items without known labels is the classic clustering pattern.