Google Cloud (GCP)
Google's cloud platform — strongest in data, AI/ML workloads, and Kubernetes-native infrastructure.
Google Cloud Platform completes the 'big three' alongside AWS and Azure. GCP has a narrower enterprise footprint than its rivals but leads in data analytics (BigQuery), machine learning infrastructure (Vertex AI, TPUs), and Kubernetes (GKE is the reference implementation). Companies with large data pipelines or ML workloads frequently reach for GCP even when their primary cloud is AWS or Azure.
Typical time to job-readiness: ~3 months.
Learning Google Cloud (GCP)
Beginner
Navigate the GCP console, deploy something to Cloud Run or App Engine, and query data in BigQuery. Create a basic Cloud Storage bucket and understand IAM basics.
Intermediate
Cloud Build for CI/CD, GKE for Kubernetes workloads, BigQuery with partitioned tables and scheduled queries, and IAM policy design.
Advanced
Pub/Sub and Dataflow for streaming pipelines, Vertex AI for ML workloads, multi-project organization design, and the Google Cloud Professional certification. Assessed through architecture case studies.
Key concepts
- BigQuery — GCP's flagship serverless data warehouse; pay per query, scales to petabytes
- Cloud Run — run containers without managing infrastructure; scales to zero when idle
- GKE (Google Kubernetes Engine) — the reference Kubernetes implementation; more mature than EKS/AKS
- IAM bindings: members, roles (Viewer/Editor/Owner + custom), and resources
- Pub/Sub — managed message queue for decoupling services and streaming data ingestion
- GCP projects — the fundamental billing and access boundary (equivalent to AWS accounts)
Common interview topics
- When would you choose GCP over AWS for a data-heavy workload
- Explain how BigQuery's pricing and performance model works
- What is Cloud Run and how does it compare to Cloud Functions
- How does IAM work in GCP
- What GCP services would you use to build a real-time data pipeline