Specialization
Cloud Computing
Cloud computing is renting someone else's data center — compute, storage, networking, and dozens of managed services — instead of buying and running your own servers. This module covers the three hyperscale providers that dominate the industry (AWS, Azure, Google Cloud), the equivalent service in each for every major need, and how to actually reason about which one — or which combination — fits a given job. DevOps — the tooling and practice of shipping to that cloud reliably — is covered as its own dedicated module right after this one.
Topic Map
Everything in this module, at a glance
In Depth
Every topic, explained — with real-world industry context
Cloud Fundamentals
Before comparing providers, you need the vocabulary and mental model of what "the cloud" actually is and the different ways you can consume it.
Basics
- What cloud computing actually is — renting infrastructure, not owning it
- IaaS, PaaS & SaaS — the three service models, with real examples
- Public, private & hybrid cloud
- Regions & availability zones
Intermediate
- The major providers — AWS, Microsoft Azure & Google Cloud (GCP)
- Smaller/specialized providers — DigitalOcean, Cloudflare, Oracle Cloud
- Pay-as-you-go pricing & how cloud billing actually works
- The shared responsibility model — what the provider secures vs. what you do
Advanced
- Multi-cloud strategy — when it's genuinely worth the complexity
- Vendor lock-in — what it costs to switch providers later
- Cloud-native design principles
- Evaluating a new cloud service before adopting it
Real-World Industry Use Case
AWS holds roughly 31% market share with 200+ managed services, Azure sits at ~23-25% driven by Microsoft 365/Active Directory integration, and GCP holds ~11-12% while growing fastest — three genuinely different bets, which is exactly why understanding the trade-offs (not just one provider) matters.
Compute Services
Compute is where your actual code runs — and every provider offers the same three shapes of it: virtual machines, containers, and serverless functions.
Basics
- Virtual machines — AWS EC2, Google Compute Engine, Azure VMs
- Choosing an instance size & type
- Serverless functions — AWS Lambda, Google Cloud Functions, Azure Functions
- When to use a VM vs. a serverless function
Intermediate
- Container-based compute — AWS Fargate, Google Cloud Run, Azure Container Apps
- Auto-scaling — adding capacity automatically under load
- Cold starts & serverless performance trade-offs
- Spot/preemptible instances for cost savings
Advanced
- Choosing compute for a specific workload — batch, real-time, bursty
- Managed Kubernetes as a compute layer (EKS, GKE, AKS) — a preview of the DevOps module
- Cost-optimizing a compute-heavy workload
- GPU/specialized compute for AI workloads
Real-World Industry Use Case
AWS wins on raw compute breadth with 500+ instance types, Google Cloud Run offers near-zero cold starts for any containerized app in any language, and Azure remains the strongest choice specifically for .NET and Windows-based workloads — the "right" compute choice genuinely depends on what you're running.
Storage Services
Cloud storage comes in a few fundamentally different shapes, each suited to a different kind of data.
Basics
- Object storage — Amazon S3, Google Cloud Storage, Azure Blob Storage
- Block storage — attached disks for a virtual machine
- File storage — shared, mountable file systems
- Uploading, retrieving & organizing objects
Intermediate
- Storage classes & lifecycle policies — hot, cool & archive tiers
- Versioning & backup strategies for object storage
- Access control on storage buckets — a common real security gap
- Static website & asset hosting on object storage
Advanced
- Cost-optimizing storage at scale (lifecycle rules, tiering automation)
- Cross-region replication for durability
- Data lakes built on object storage
- Egress costs — the hidden cost of moving data out of a cloud
Real-World Industry Use Case
A misconfigured, publicly-readable S3 bucket is one of the most common real-world causes of major data breaches — which is exactly why storage access control is treated as a core topic here, not an afterthought.
Networking & Content Delivery
Networking is how your cloud resources talk to each other and to the internet — and how content gets delivered quickly to users anywhere in the world.
Basics
- Virtual networks — AWS VPC, Azure Virtual Network, GCP VPC
- Subnets — public vs. private
- Load balancers — distributing traffic across servers
- DNS in the cloud — AWS Route 53 & equivalents
Intermediate
- CDNs — CloudFront, Cloud CDN & Azure CDN for fast global delivery
- Security groups & firewalls
- API gateways as a networking/traffic layer
- Private connectivity between services (VPC peering)
Advanced
- Global load balancing & multi-region traffic routing
- Google's private global network as a competitive differentiator
- Zero-trust networking principles in the cloud
- Diagnosing latency & connectivity issues across regions
Real-World Industry Use Case
GCP's private subsea fiber network routes traffic away from the public internet specifically to deliver more consistent global latency — a genuine, measurable differentiator companies factor into their provider choice, not just marketing.
Managed Databases & Data Warehousing
Every major provider offers a managed version of the databases covered in the Databases & SQL module — someone else handles patching, backups & scaling for you.
Basics
- Managed relational databases — Amazon RDS, Cloud SQL, Azure SQL Database
- Managed NoSQL — DynamoDB, Firestore, Azure Cosmos DB
- Why "managed" saves real operational work
- Connecting an app to a managed database
Intermediate
- Automated backups, patching & failover in managed databases
- Read replicas as a managed-service feature
- Managed caching — ElastiCache, Memorystore
- Choosing a managed database tier for a given workload
Advanced
- Data warehousing — Amazon Redshift, Google BigQuery, Azure Synapse
- Serverless databases (Aurora Serverless) & scale-to-zero economics
- Cross-service data pipelines feeding a warehouse
- Cost modeling for a managed data platform at scale
Real-World Industry Use Case
BigQuery is widely considered the strongest managed data warehouse on a cost-per-query basis for large-scale analytics — a genuine reason some companies choose GCP as their primary cloud specifically because their data strategy centers on it.
Identity, Security & Cost Management
Cloud security is a shared responsibility, and cloud cost is easy to lose control of — both require dedicated, deliberate practice, not defaults.
Basics
- IAM — Identity and Access Management, and the principle of least privilege
- Users, roles & permission policies
- Reading a cloud bill & understanding line items
- Setting up budget alerts
Intermediate
- Service accounts & machine-to-machine access
- Encryption at rest & in transit, provider-managed vs. customer-managed keys
- Cost allocation tags & tracking spend by project/team
- Compliance basics — SOC 2, GDPR-relevant cloud configuration
Advanced
- Reserved instances & savings plans for predictable workloads
- Auditing IAM policies for over-permissioned access
- Multi-account/multi-project security architecture
- FinOps — the discipline of ongoing cloud cost optimization
Real-World Industry Use Case
An over-permissioned IAM role or an exposed access key is one of the most common real-world cloud security incidents — and an unmonitored auto-scaling group is one of the most common ways a company gets an unexpectedly enormous cloud bill.
Choosing & Comparing Providers
By this point you know the service categories — this closes the loop by comparing the providers directly, so you can make a real, defensible choice for a real project.
Basics
- AWS — broadest service catalog, market leader, most job postings
- Azure — best for .NET/Windows shops & Microsoft-integrated companies
- Google Cloud — best for data/analytics & Kubernetes-native teams
- Reading a provider's pricing page without getting overwhelmed
Intermediate
- AI/ML capabilities as a 2026 differentiator — Azure/OpenAI, GCP TPUs & BigQuery ML, AWS SageMaker
- Comparing egress costs & architectural lock-in across providers
- Smaller providers for simpler needs — DigitalOcean, Vercel, Cloudflare
- Matching a provider to a team's existing skills & stack
Advanced
- Running a real cost/feature comparison for a hypothetical project
- Designing for portability to reduce future lock-in risk
- When multi-cloud is a genuine architecture decision vs. resume-driven complexity
- Staying current as provider capabilities shift year to year
Real-World Industry Use Case
AI/ML capability is now considered the single biggest differentiator between providers — Azure for OpenAI/GPT access, GCP for TPUs and BigQuery ML, AWS for the broadest GPU selection and SageMaker ecosystem — proof that "which cloud" is a strategic decision, not just a pricing one.
Ready to learn this — live?
See how this module fits into the complete curriculum.