Study CLF-C02 Analytics, AI/ML, and Integration: key concepts, common traps, and exam decision cues.
This lesson covers the part of CLF-C02 where AWS throws many managed-service names at you and expects you to stay calm. The exam is still broad, not deep. The real task is to recognize which service family fits the job: analytics, AI/ML, messaging, event routing, developer tooling, end-user computing, frontend, or IoT.
Publish-subscribe: Messaging pattern where one publisher sends a message to many subscribers.
Event routing: Delivering events between services or systems based on rules rather than simple queueing.
AWS wants you to classify:
| Need | Strongest first fit | Why |
|---|---|---|
| queue work for later processing | Amazon SQS | This is the asynchronous queue lane. |
| publish notifications or fan out messages | Amazon SNS | This is the publish-subscribe lane. |
| route events between AWS services and systems | Amazon EventBridge | This is the event-routing lane. |
| Need | Strongest first fit | Why |
|---|---|---|
| query data in S3 | Amazon Athena | This is the managed query lane for data in S3. |
| prepare or move data for analytics | AWS Glue | This is the data integration and ETL lane. |
| stream and process real-time data | Amazon Kinesis | This is the streaming analytics lane. |
| business intelligence dashboards and visualization | Amazon QuickSight | This is the visualization lane. |
| broader machine learning model development | Amazon SageMaker | This is the general ML platform lane. |
| conversational bots | Amazon Lex | This is the conversational AI lane. |
| search, vision, or voice use cases | Kendra, Rekognition, Polly and related managed services | These are narrower managed AI use-case services. |
CLF-C02 also expects broad awareness of:
The goal is not mastery. It is category recognition.
This is one of the most common fundamentals traps.
| If the requirement is mainly about… | Think first about… |
|---|---|
| queued asynchronous work | SQS |
| sending one message to many subscribers | SNS |
| routing events across services and systems | EventBridge |
If the question is about “messages wait until processed,” SQS is usually stronger than SNS. If it is about “notify many subscribers,” SNS is usually stronger. If it is about rule-based event movement between systems, EventBridge is usually stronger.
| Trap | Better thinking |
|---|---|
| “Every integration question is probably EventBridge.” | First ask whether the real need is queueing, fan-out, or event routing. |
| “Every AI/ML question means SageMaker.” | Some stems are really about a narrower managed AI service such as Lex or Rekognition. |
| “QuickSight is basically storage.” | QuickSight is a visualization and BI service, not storage. |
| “Developer tools and runtime services are the same category.” | Build-and-deploy services are different from services that run application logic. |
A company wants to send one notification to many subscribers whenever a new event occurs. The requirement emphasizes message fan-out, not queued work.
Which service is strongest first?
Correct answer: 2
Why: The requirement is publish-subscribe notification fan-out, which is the SNS lane. SQS is stronger for queued asynchronous work.