AWS Solutions Architect & DevOps Course: SAA-C03 + DOP-C02

Identity is the first thing an architect designs and the first thing an auditor checks. We start here rather than with virtual machines, because every later decision inherits from it.

• The global infrastructure: regions, availability zones and edge locations
• IAM users, groups, roles, policies and permission boundaries
• AWS Organizations, service control policies and multi account strategy
• Identity Center for workforce access, and why long lived keys are a liability
• Billing, budgets, cost allocation tags and the alerts that prevent surprises
• Hands on: a multi account structure with guardrails, built properly

Networking carries the most weight in architecture interviews because it is the layer that is painful and expensive to change later.

• VPC design, subnetting, route tables, internet and NAT gateways
• Security groups against network ACLs, and when each applies
• VPC peering, Transit Gateway and PrivateLink for service to service access
• Hybrid connectivity with Site to Site VPN and Direct Connect
• Route 53 routing policies, health checks and failover
• CloudFront and edge delivery
• Hands on: a secure multi availability zone VPC built from an empty account

The exam hands you a workload and asks which service fits. This module builds the judgement to answer that quickly and defend it.

• EC2 instance families, purchasing models and placement groups
• EBS volume types, snapshots and instance store trade offs
• S3 storage classes, lifecycle policies, versioning, encryption and replication
• RDS and Aurora: high availability, read replicas, backup and restore
• DynamoDB data modelling, capacity modes, global tables and DAX
• Choosing between relational, key value, cache and object storage under constraints
• Hands on: design and build the full data layer for a web application

The revised architect exam puts more weight on resilient design and cost optimisation, so this module carries a full weekend rather than a passing mention.

• Elastic Load Balancing types and target group design
• Auto Scaling policies, health checks and self healing behaviour
• Multi availability zone and multi region patterns, with the cost of each stated honestly
• Backup, recovery objectives and designing to a real RTO and RPO
• The Well Architected Framework applied as a working method, not a slide
• Right sizing, savings plans, spot capacity and storage tiering
• Hands on: a self healing three tier architecture with a cost model attached

Serverless and decoupling questions run through the whole exam, and event driven design is what modern AWS estates actually look like.

• Lambda: execution model, concurrency, cold starts and limits
• API Gateway, authorisers, throttling and usage plans
• SQS, SNS and EventBridge for decoupled, event driven systems
• Step Functions for orchestration and long running workflows
• Idempotency, retries, dead letter queues and poison message handling
• Hands on: a serverless API and an event driven pipeline, end to end

A full weekend of exam conditions rather than revision slides, because the architect exam rewards fast, confident elimination of near miss answers.

• All exam domains reviewed in blueprint proportion
• Scenario drills: reading a question for the constraint that decides the answer
• A full length domain scored mock plus a score report review
• Exam booking, pacing and question strategy
• A revision plan for the two weeks after this weekend

The professional exam assumes you automate rather than click. This module gives that to you from nothing, including for career switchers.

• Linux essentials that DevOps work actually requires
• Python from the beginning: structure, error handling, packaging
• The AWS CLI and boto3 for scripted operations
• Writing automation that is safe to run twice
• Hands on: an automation toolkit for the tasks you would otherwise repeat by hand

SDLC automation is the largest domain on the professional exam. This is where most of the marks and most of the real world value sit.

• Git branching strategies teams genuinely use, and their failure modes
• CodeCommit, CodeBuild, CodeDeploy and CodePipeline as a working chain
• Jenkins pipelines and shared libraries for organisations already invested in it
• Deployment strategies: in place, blue green, canary and linear
• Artefact management, versioning and promotion between environments
• Hands on: a commit to production pipeline with approvals and automated rollback

The exam expects CloudFormation fluency and the market expects Terraform, so you learn both and understand where CDK fits between them.

• CloudFormation templates, nested stacks, change sets and drift detection
• StackSets for multi account and multi region deployment
• AWS CDK for teams that would rather express infrastructure in code
• Terraform: providers, state, modules, workspaces and remote backends
• Multi environment patterns and policy as code guardrails
• Hands on: an entire environment created from code in a single apply

Containers are the default delivery target on AWS now, and operating EKS is the skill that separates senior candidates from the rest.

• Docker images, multi stage builds and image hygiene
• Elastic Container Registry, scanning and lifecycle policies
• ECS and Fargate: task definitions, services and when serverless containers win
• EKS core objects: deployments, services, ingress, config and secrets
• Node groups, autoscaling, IRSA and cluster upgrades
• Hands on: containerise a microservice and run it on EKS behind ingress

Monitoring, incident response and security automation together carry a large share of the professional exam, and they are what makes being on call survivable.

• CloudWatch metrics, logs, alarms, dashboards and Logs Insights queries
• X-Ray tracing, service maps and finding latency in a distributed system
• EventBridge driven automated remediation and self healing responses
• Config, GuardDuty, Security Hub and compliance as code
• Secrets Manager, Parameter Store and eliminating stored credentials
• Incident response, runbooks and blameless postmortems
• The full professional exam outline in blueprint proportion, plus a domain scored mock

The layer nobody else teaches. Infrastructure code, runbooks and incident notes are already being written with AI assistance in the teams that are hiring, and using it well is now a hiring signal.

• Generating and reviewing Terraform and pipeline code with Claude, and catching what it gets wrong
• Runbook copilots and AI assisted incident response wired to CloudWatch alarms
• Automated cost and security review passes across an account
• Claude Code for repository wide refactors and migrations
• Capstone: ship a monitored, infrastructure as code managed service and present the architecture
• Mock interview defending your design and operational choices