Skill Roadmap

AWS

AWS has a lot of services — you don't need to master all of them. This roadmap covers an opinionated path through the ones that matter most: access control and networking, compute, storage, core managed services, databases and containers, and serverless with Lambda.

What skills does an AWS engineer need?

It depends how far into the ecosystem you want to go

Before touching any single service, it helps to understand cloud computing itself — the difference between IaaS, PaaS, and SaaS, public vs private vs hybrid cloud, AWS's global infrastructure of regions and availability zones, and the Shared Responsibility Model that defines what AWS secures versus what you do.

From there, almost every path runs through the same three essentials: IAM for access control, VPC for networking, and EC2 for compute. After that it branches by direction — storage and content delivery, managed databases and containers, or going fully serverless with Lambda.

The best way to learn is to build a simple app and use these services to deploy it end to end, rather than studying each one in isolation. A working grasp of the Well-Architected Framework rounds things out once you're designing real systems, not just standing up individual resources.

The AWS Roadmap

Work through these in order, then deploy a small app using what you've learned

STEP 1
AWS logo

Cloud Fundamentals

The concepts and vocabulary every AWS service builds on.

What is Cloud Computing? IaaS vs PaaS vs SaaS Public vs Private vs Hybrid Cloud AWS Global Infrastructure Shared Responsibility Model Well-Architected Framework Introduction to AWS
STEP 2

Essentials — IAM, VPC & EC2

Access control, networking, and compute — the three services almost everything else depends on.

IAM — access control
Policies Users / User Groups Roles Identity-based Resource-based Instance Profiles Assuming Roles
VPC — network layer
CIDR Blocks Subnets Route Tables Security Groups Internet Gateway NAT Gateway Private Subnet Public Subnet
EC2 — compute
Instance Types CPU Credits Storage / Volumes Keypairs Elastic IP User Data Scripts Purchasing Options AMIs Launch Templates Auto-Scaling Groups Scaling Policies Elastic Load Balancers
STEP 3

Core Services — S3, SES, Route53, CloudWatch & CloudFront

Storage, email, DNS, monitoring, and content delivery — the next layer to learn.

S3 — storage
Buckets / Objects Bucket / Object Lifecycle Standard S3-IA Glacier
SES — email
Sandbox / Sending Limits Identity Verification DKIM Setup Feedback Handling Configuration Sets Sender Reputation Dedicated IP
Route53 — DNS
Hosted Zones Routing Policies Health Checks
CloudWatch — monitoring
Metrics Events Logs
CloudFront — CDN
Distributions Policies Invalidations
STEP 4

Databases, Caching & Containers

Managed data stores and container orchestration for running real workloads.

RDS — managed database
DB Instances Storage Types Backup / Restore General Purpose Provisioned IOPS Magnetic
DynamoDB — NoSQL storage
Tables / Items / Attributes Primary Keys / Secondary Indexes Data Modeling Streams Capacity Settings Limits DynamoDB Local
ElastiCache — Redis / Memcached
Quotas
ECS & EKS — containers
ECR Clusters / Container Agents Tasks Services Launch Config / Auto-Scaling Groups Fargate EKS
STEP 5

Serverless

Running code without managing servers at all.

Lambda
Creating / Invoking Functions Layers Custom Runtimes Versioning / Aliases EventBridge / Scheduled Execution Cold Start and Limitations API Gateway Lambda@Edge
Also worth knowing
ECS Fargate

GitHub Projects

Real, buildable projects to put on your own GitHub

Frequently Asked Questions

Common questions from people starting out with AWS

Do I need to learn all of AWS?

No — AWS has hundreds of services, and no single engineer uses them all. Focus on IAM, VPC, and EC2 first, then branch into the services that match the kind of system you're building.

Is AWS free to use?

AWS offers a free tier covering limited usage of services like EC2, S3, and Lambda for a set period or amount, which is enough to practice most of this roadmap. Usage beyond those limits is billed, so it's worth watching your resources closely while learning.

Do I need to know networking before AWS?

Not in depth, but a basic understanding of IP addressing, subnets, and how routing works will make VPC concepts like CIDR blocks, route tables, and gateways click much faster.

Should I learn AWS or another cloud provider first?

AWS holds the largest market share and the broadest job demand, which is why it's a common starting point. That said, the core concepts — IAM, networking, compute, storage — transfer well to Azure or Google Cloud once you understand them on one platform.

What's the difference between EC2 and Lambda?

EC2 gives you a persistent virtual server you manage and pay for continuously. Lambda runs your code only when triggered and you're billed per invocation, trading control over the environment for less operational overhead.

How do I avoid a surprise AWS bill?

Set up billing alerts and a budget in the AWS console early, shut down or terminate resources like EC2 instances and RDS databases when you're not using them, and keep an eye on data transfer and storage costs, which add up quietly.

Track complete

From IAM and networking through compute, storage, databases, and serverless — that's an opinionated path through the AWS services most engineers actually use. Deploy a real app with them, then let the direction you enjoy pull you deeper into containers, data, or serverless.

Where next?

Keep exploring by domain or drill into a single skill