Skip to content
Devgo

AWS (Amazon Web Services)

The largest cloud computing platform, offering 200+ services: compute (EC2), storage (S3), databases (RDS, DynamoDB), and more. AWS is the backbone of much of the internet. Powerful but complex, with a steep learning curve and pricing that can surprise you. Used by Netflix, Airbnb, Coinbase.

Created by

Amazon · Launched AWS in 2006, led by Andy Jassy

Website
When to use
  • When you need a full cloud platform with many services.
  • Global-scale apps.
  • Companies wanting managed everything.
When not to
  • Simple sites (use Vercel/Netlify).
  • When you need predictable pricing (DigitalOcean).
  • Small teams without cloud expertise.

Definition

AWS is Amazon's cloud platform with 200+ services across compute (EC2, Lambda), storage (S3, EBS), databases (RDS, , Aurora), networking (VPC, CloudFront), AI (SageMaker, Bedrock), and more. Regions and Availability Zones for global . Pay-as-you-go pricing.

Why it exists

To provide on-demand computing infrastructure so companies don't build data centers — scaling globally without upfront hardware costs, with managed services for common needs.

How it works

Create an AWS account. Launch services via console, CLI, or IaC (Terraform, CloudFormation). EC2: virtual machines. S3: . Lambda: functions. RDS: managed databases. VPC: private networks. IAM: access control. Pay per usage (hours, GB, requests).

Scaling limits

Complexity (200+ services, deep knowledge needed). Pricing is hard to predict (many dimensions). Vendor lock-in. Service limits (quotas). Multi-region complexity. Free tier is limited.

Cost shape

Pay-as-you-go. Free tier (limited, 12 months). Can get expensive fast (data transfer, unused resources). Reserved Instances/Savings Plans reduce cost. Many startups get credits.

Alternatives

  • GCP (Google Cloud) — Strong in AI/data, smaller market share.
  • Azure — Strong enterprise/Windows integration.
  • DigitalOcean/Vultr — Simpler, cheaper for small apps, fewer services.

Failure modes

  • Unexpected bills (data transfer, unused resources).
  • Security misconfiguration (public S3 buckets).
  • Service complexity.
  • Vendor lock-in.
  • Outages in a region.

, ,