Google Cloud Platform groups its services into a handful of logical categories, compute, storage, networking, databases, and AI, so picking the right tool mostly comes down to matching the category to your actual workload rather than learning the entire catalog.
Compute and Storage
- Compute Engine: custom VMs with your choice of CPU, memory, and OS
- Google Kubernetes Engine: managed Kubernetes for containerized apps
- Cloud Run: fully managed containers that scale to zero, now the recommended default for new serverless projects over App Engine
- Cloud Functions and App Engine: event-driven code and (mostly legacy at this point) fully managed web app hosting
- Cloud Storage, Persistent Disk, and Filestore for object, block, and file storage respectively
Full detail on compute specifically is in Google Cloud computing services; storage classes and pricing are in Google Cloud storage services.
Networking and Databases
VPCs isolate and segment your network, Cloud Load Balancing distributes traffic across instances and regions, and Cloud CDN caches content at the edge. For data, Cloud SQL runs managed MySQL, PostgreSQL, or SQL Server, Cloud Spanner offers a globally distributed database with ACID transactions, and Firestore is a serverless document database built for web and mobile apps.
AI and Migration
The Gemini Enterprise Agent Platform (Google’s rebranded, expanded successor to Vertex AI as of Google Cloud Next 2026) covers training, deploying, and monitoring custom machine learning models end to end, and AutoML builds models with minimal coding on top of it, see Google Cloud AI services for the full picture. For moving existing workloads in, the Google Cloud Migration Service replicates live databases with minimal downtime, and Transfer Appliances handle large offline transfers, detailed in Google Cloud migration services.
Security and Support
Cloud IAM, VPC firewall rules, Cloud Armor for DDoS protection, and Customer Managed Encryption Keys cover the core security surface, full detail in Google Cloud security services. For hands-on help, Google Cloud consulting services covers architecture and strategy, Google Cloud professional services handles large rollouts and performance tuning, and Google Cloud managed services takes on routine operations once you are live.
Picking a Starting Point
For a team new to the platform, the fastest path to a working mental model isn’t reading the full catalog, it’s picking the category matching whatever you’re building right now (a container-based app, a data pipeline, an ML model) and going one level deep into that category’s dedicated guide above, then coming back to the rest as an actual need for it shows up. Most teams never touch more than three or four of these categories in a given project.