Executive Summary
In the previous lessons, you learned what cloud computing is, how cloud providers differ, and how responsibilities are shared between providers and customers.
Related Lesson: What is Cloud Computing ?
Related Lesson: Major Cloud providers offerings and differences
The next step is understanding how cloud resources are actually created.
Whether deploying a virtual machine, creating a storage bucket, provisioning a Kubernetes cluster, or deploying a database, every cloud resource is created through one of several methods.
Modern cloud platforms support:
- Management Consoles
- Command Line Interfaces (CLI)
- Software Development Kits (SDK)
- APIs
- Infrastructure as Code (IaC)
More recently, AI and Agentic AI platforms have started generating and deploying infrastructure through these same interfaces.
Understanding these resource creation methods is essential because they influence automation, governance, consistency, scalability, and operational excellence across cloud environments.
Why Resource Creation Matters
Imagine two engineers deploying the same virtual machine.
The first engineer manually clicks through the cloud console and creates the resource.
The second engineer deploys the same resource using Infrastructure as Code through a CI/CD pipeline.
The outcome appears identical.
The operating model is not.
One approach is manual and difficult to reproduce consistently.
The other is automated, version-controlled, auditable, and scalable.
As organizations grow, the method used to create resources often becomes more important than the resource itself.
Related Lesson: Building Blocks of Multi-Cloud Computing
This is one reason automation, Infrastructure as Code, and platform engineering have become foundational practices in modern cloud environments.
Learning Objectives
After completing this lesson, you should be able to:
- Understand how cloud resources are created.
- Explain the differences between Console, CLI, SDK, API, and Infrastructure as Code.
- Understand imperative and declarative provisioning models.
- Evaluate resource creation methods from both engineer and architect perspectives.
- Understand why Infrastructure as Code became the enterprise standard.
- Recognize how AI and Agentic AI are changing cloud operations.
How Cloud Resources Are Created
Every cloud resource ultimately starts with a request.
That request may come from a user, an application, an automation pipeline, Infrastructure as Code, or an AI agent.
Regardless of the method used, cloud platforms process the request through APIs before creating resources.
The figure below illustrates this concept.
This is one of the most important concepts in cloud computing.
Everything eventually becomes an API request.
The console, CLI, SDK, Infrastructure as Code tools, and AI agents are simply different ways of interacting with those APIs.
The Evolution of Resource Provisioning
Cloud computing transformed how infrastructure is provisioned.
The progression below shows how infrastructure management evolved over time.
Each step increased:
- Automation
- Consistency
- Scalability
- Governance
while reducing manual effort.
Modern enterprises increasingly favor Infrastructure as Code and automation-driven approaches because they provide repeatable and auditable deployment processes.
The Five Ways Cloud Resources Are Created
Cloud providers expose multiple interfaces for creating and managing resources.
Each method serves a different purpose.
Console
The cloud console is the web-based graphical interface provided by cloud vendors.
Examples include:
- AWS Management Console
- Azure Portal
- Google Cloud Console
- OCI Console
- IBM Cloud Console
The console is often the easiest place for beginners to start.
It provides visual workflows and requires minimal setup.
Typical use cases:
- Learning cloud services
- Exploring configurations
- Small deployments
- Troubleshooting
However, manual console-based deployments can become difficult to scale consistently across large environments.
Command Line Interface (CLI)
A Command Line Interface allows engineers to create and manage resources through terminal commands.
Examples include:
- AWS CLI
- Azure CLI
- gcloud CLI
- OCI CLI
- IBM Cloud CLI
CLI tools are commonly used for:
- Automation scripts
- Administration tasks
- Bulk operations
- Troubleshooting
Many engineers prefer CLI tools because they are faster and easier to automate than graphical interfaces.
Software Development Kits (SDK)
SDKs allow developers to interact with cloud services directly from application code.
Common SDK languages include:
- Python
- Java
- JavaScript
- Go
- .NET
Typical use cases include:
- Application integration
- Dynamic infrastructure interactions
- Automation tooling
- Custom platforms
SDKs are widely used when applications need to interact directly with cloud services.
APIs
Every cloud platform exposes APIs that allow resources to be created programmatically.
In reality:
Console
↓
CLI
↓
SDK
↓
APIAll roads eventually lead to APIs.
APIs provide the foundation for:
- Automation
- Infrastructure as Code
- Platform engineering
- Cloud-native applications
- AI-driven operations
Understanding this relationship helps explain how cloud platforms operate behind the scenes.
Infrastructure as Code (IaC)
Infrastructure as Code represents the modern enterprise standard for resource provisioning.
Instead of manually creating resources, engineers define infrastructure in code.
Examples include:
- Terraform
- OpenTofu
- AWS CloudFormation
- Azure Bicep
- Pulumi
Infrastructure definitions can then be:
- Version controlled
- Reviewed
- Tested
- Automated
- Reused
This approach significantly improves consistency and governance across environments.
Resource Creation Methods Compared
The comparison below highlights common strengths of each provisioning approach.
| Method | Best For | Automation Level |
|---|---|---|
| Console | Learning and small deployments | Low |
| CLI | Administration and scripting | Medium |
| SDK | Application integration | Medium |
| API | Platform integrations | High |
| Infrastructure as Code | Enterprise provisioning | Very High |
No single method is universally better.
Most organizations use multiple methods depending on the workload and operational requirements.
Imperative vs Declarative Resource Creation
One of the most important concepts in automation is understanding the difference between imperative and declarative approaches.
The figure below summarizes the difference.
Imperative
Create VPC
Create Subnet
Create Route Table
Create VMDeclarative
Desired State:
One VPC
One Subnet
One VMInfrastructure as Code tools largely embrace declarative approaches because they are easier to automate, validate, and govern at scale.
Resource Creation: Engineer and Architect Perspective
Engineers and architects often evaluate provisioning methods differently.
The comparison below highlights these perspectives.
As environments grow, architects typically prioritize approaches that improve repeatability, auditability, and operational consistency.
Resource Provisioning Across major Cloud Providers
Although cloud providers use different service names and management tools, the underlying provisioning model is remarkably similar.
Every cloud platform ultimately exposes APIs that can be accessed through consoles, CLIs, SDKs, Infrastructure as Code tools, and increasingly through AI-assisted workflows.
The comparison below shows how common provisioning methods map across major cloud providers.
While the tools differ, the architecture principle remains the same:
Every cloud resource is ultimately created through an API.
Resource Provisioning in Multi-Cloud Environments
As organizations adopt multiple cloud providers, resource provisioning becomes less about individual tools and more about standardization.
Without common deployment patterns, teams often end up maintaining:
- Different deployment processes
- Different security controls
- Different tagging standards
- Different operational procedures
The figure below illustrates a common multi-cloud provisioning approach.
The goal is not to make every cloud identical.
The goal is to establish consistent provisioning, governance, security, and operational practices across providers.
Common enterprise approaches include:
- Infrastructure as Code as the primary provisioning method
- Standardized tagging strategies
- Shared security baselines
- Centralized governance policies
- Common CI/CD deployment pipelines
These practices help reduce operational complexity as cloud environments expand.
Why Infrastructure as Code Became the Enterprise Standard
Most organizations begin their cloud journey using management consoles.
As environments grow, manual provisioning becomes difficult to govern consistently.
The figure below illustrates why Infrastructure as Code became the preferred provisioning approach for modern enterprises.
Infrastructure as Code enables organizations to:
- Reuse deployment patterns
- Standardize configurations
- Improve auditability
- Reduce configuration drift
- Support automated deployments
This is one reason Terraform, OpenTofu, CloudFormation, Bicep, and Pulumi have become foundational technologies in cloud operations.
Platform Engineering and Self-Service Provisioning
As cloud adoption grows, many organizations move beyond simple Infrastructure as Code deployments and begin building internal platforms.
The figure below illustrates a common platform engineering workflow.
Instead of every team creating infrastructure differently, platform engineering promotes:
- Standardized deployment patterns
- Security by default
- Faster provisioning
- Operational consistency
- Reduced cognitive load for developers
This model is becoming increasingly common in large enterprises operating across multiple cloud providers.
Resource Provisioning with AI and Agentic AI
Artificial Intelligence is beginning to influence how infrastructure is designed, generated, reviewed, and deployed.
Rather than replacing cloud engineers, AI is helping teams accelerate provisioning workflows.
The figure below illustrates a common AI-assisted provisioning pattern.
Current enterprise use cases include:
- Generating Terraform templates
- Creating deployment scripts
- Explaining infrastructure errors
- Reviewing architecture patterns
- Producing operational documentation
Emerging capabilities include:
- Policy-aware infrastructure generation
- Automated remediation workflows
- Agent-driven provisioning
- Self-service infrastructure assistants
Organizations adopting AI-assisted provisioning typically maintain approval workflows and governance controls to ensure quality and compliance.
Well-Architected Multi-Cloud Strategy
Resource provisioning is not simply a deployment activity. It directly affects security, reliability, governance, performance, and cost management.
The figure below illustrates how provisioning strategies should be evaluated through a Well-Architected lens.
Multi-Cloud Considerations
| Area | Multi-Cloud Focus |
| Operational Excellence | Standardized deployment pipelines |
| Security | Consistent policy enforcement |
| Reliability | Repeatable recovery processes |
| Performance | Consistent architecture standards |
| Cost Optimization | Centralized governance and visibility |
Organizations that standardize provisioning processes typically achieve stronger governance and operational outcomes than those relying heavily on manual deployments.
Enterprise Best Practices for Resource Provisioning
As cloud environments grow, resource provisioning becomes less about creating infrastructure and more about creating consistent, secure, and repeatable operating models.
The figure below summarizes the foundational practices commonly found in mature multi-cloud environments.
These capabilities provide the foundation for scalable cloud operations across both single-cloud and multi-cloud environments.
The following checklist highlights common enterprise provisioning practices.
Identity and Access Management
✓ Standardize authentication and authorization models
✓ Use centralized identity providers where possible
✓ Apply least-privilege access principles
Resource Standards
✓ Establish consistent naming conventions
✓ Implement organization-wide tagging standards
✓ Define approved deployment templates
Security and Compliance
✓ Apply security baselines before deployment
✓ Validate configurations through automated policy checks
✓ Integrate compliance controls into deployment pipelines
Automation and Infrastructure as Code
✓ Treat infrastructure definitions as code
✓ Store infrastructure in version control systems
✓ Use automated deployment pipelines
Monitoring and Operations
✓ Deploy monitoring by default
✓ Standardize logging and alerting configurations
✓ Ensure resources are discoverable and traceable
Multi-Cloud Governance
✓ Use common deployment standards across providers
✓ Centralize governance and policy management
✓ Maintain visibility into costs, security, and compliance
Organizations that establish these practices early typically experience fewer operational challenges as cloud adoption expands.
Common Mistakes and Misconceptions
Many cloud provisioning challenges are caused by process issues rather than technology limitations.
The comparison below highlights some of the most common misconceptions.
| Misconception | Reality |
|---|---|
| The cloud console is sufficient for enterprise deployments | Large environments require automation and governance |
| Infrastructure as Code is only for large organizations | Even small environments benefit from consistency and version control |
| Multi-cloud means using different tools for every provider | Standardization reduces operational complexity |
| AI can safely deploy infrastructure without oversight | Human review and governance remain essential |
| Automation removes the need for architecture decisions | Automation accelerates implementation but does not replace architecture |
Understanding these misconceptions helps organizations avoid many common operational challenges.
Architect’s Notebook
The figure below summarizes practical lessons commonly observed across enterprise cloud environments.
Key Takeaways
The figure below summarizes the evolution of cloud resource provisioning.
Key lessons from this chapter:
- Every cloud resource is ultimately created through an API.
- Consoles, CLIs, SDKs, and Infrastructure as Code are different interfaces to the same cloud platform.
- Infrastructure as Code has become the preferred enterprise provisioning model.
- Multi-cloud environments require standardized deployment practices.
- Platform engineering builds on Infrastructure as Code to provide self-service capabilities.
- AI and Agentic AI are accelerating provisioning workflows but do not remove governance requirements.
- Resource provisioning is a foundational capability that influences security, reliability, governance, and operational excellence.
What’s Next
Cloud Identity and Access Management (IAM) Explained Across major Cloud Providers
Every cloud resource ultimately depends on identity.
Before users, applications, automation pipelines, or AI agents can create or access resources, they must first be authenticated and authorized.
In the next lesson, you will learn:
- Authentication and authorization fundamentals
- Users, groups, roles, and permissions
- Single Sign-On (SSO)
- Federation and identity providers
- Multi-cloud identity strategies
- Identity considerations for automation and AI-driven operations
Identity is often considered the most important building block in cloud architecture because nearly every security, governance, and operational process depends on it.
