You don't need to learn all three. Pick one, go deep, and understand that the others exist as "the same concepts, different names."
| Concept | AWS | GCP | Azure |
|---|---|---|---|
| Object storage | S3 | Cloud Storage | ADLS |
| Serverless SQL over files | Athena | BigQuery | Synapse |
| Managed orchestration | MWAA (managed Airflow) | Composer | Data Factory |
| Data catalogue | Glue | Data Catalog | Purview |
| Notebooks | SageMaker | Colab / Vertex AI | Azure ML Notebooks |
<aside> 💡
Recommendation: if you're learning from scratch, start with GCP. BigQuery's free sandbox and Colab give you a genuinely free path to real practice without a credit card. AWS is the most commonly required in job postings, so circle back to it once you're comfortable with the concepts.
</aside>
| Service | Actually free? |
|---|---|
| AWS S3 | Yes: 5GB always-free storage |
| AWS Glue | Yes: the first 1 million catalogue requests a month |
| AWS Athena | No. There's no free query tier at all. Every query costs at least $0.00005 (a 10MB minimum billed), even for a trivial test. New accounts (post-July 2025) get up to $200 in credits that can offset this while learning. |
| GCP BigQuery sandbox | Yes: a genuinely free tier with generous limits, no credit card required |
| Google Colab | Yes: a free tier with usable compute |
<aside> ⚠️
The Athena catch trips people up constantly. A "free tier" doesn't mean every service within it is free, so check the specific service rather than assuming the account tier covers it.
</aside>