Skip to Content
Cloud Engineering Academy
14. Advanced Banking Architecture

Module 14: Advanced Banking Architecture

Apply everything from earlier modules to real banking use cases, from customer onboarding and payments to fraud signals, reconciliation and regulatory reporting. Every architecture follows the same 15-section template so you practise the full reasoning a solutions architect presents to a design authority.

SAP-C03SAA-C03SCS-C03

Assumes: .NET 10 LTS · AWS SDK for .NET v4 · Region us-east-1. Verified against official sources on 2026-11-18.

Apply everything from earlier modules to real banking use cases, from customer onboarding and payments to fraud signals, reconciliation and regulatory reporting. Every architecture follows the same 15-section template so you practise the full reasoning a solutions architect presents to a design authority.

Estimated timeAbout 20 hours (architecture studies 14 h · design assignments 6 h)
Priority for youHigh — brings together priorities 1–4 (deployment, APIs, security, pipelines)
LevelAdvanced
CertificationsSAP-C03 (primary) · SAA-C03 · SCS-C03
Cost profile🟢 mostly design-only · 🟡 optional small proofs of concept

Why this matters in banking

Knowing individual services is not the same as designing a payment system that a bank's architecture board, security team and risk function will approve. Real designs must state what happens when a sanctions-screening provider times out, how a duplicate payment is prevented, which data crosses which trust boundary, and what it will cost at peak. This module trains that end-to-end judgement on scenarios you will recognise from your own work.

Nothing here is automatically compliant

These are learning architectures using fictional data. KYC/AML, sanctions screening, PSD2/Open Banking, PCI DSS, UK GDPR and operational-resilience obligations all need review by your bank's legal, compliance, risk and security teams. Each architecture lists its compliance touchpoints so you know where that review is required.

From .NET/Azure to AWS

You know (Microsoft/Azure)AWS equivalentWhat's different
Azure reference architecturesAWS Architecture Center, Well-Architected FSI Lens, AWS Solutions Library guidanceGuidance repos often ship deployable IaC samples
API Management in front of App ServiceAPI Gateway + ECS/Fargate or Lambda via VPC LinkPrivate integrations keep back ends out of the internet
Service Bus + Durable Functions for workflowsEventBridge + SQS + Step FunctionsSagas with explicit compensation states
Azure SQL ledger tablesAurora/RDS with double-entry design, DynamoDB for idempotencyIntegrity is designed in; plus KMS signing for non-repudiation
Stream AnalyticsKinesis Data Streams + Lambda, Managed Service for Apache FlinkFraud rules run on streams; ML scoring via SageMaker/Bedrock
Azure Storage + Data Factory for reportingS3 + Glue + Athena (+ Redshift)Regulatory reports are produced from a governed data lake
Azure Front Door + WAFCloudFront + AWS WAF + ShieldEdge protection for public digital-banking channels

Learning objectives

By the end of this module you can:

  • Turn a banking business requirement into functional and non-functional requirements with measurable targets.
  • Produce a complete design using the 15-section template, including rejected alternatives and failure scenarios.
  • Design payment initiation, validation, sanctions-screening hooks, settlement and idempotency end to end.
  • Integrate fraud signals without depending on services closed to new customers.
  • Design reconciliation, regulatory reporting and audit architectures with traceable evidence.
  • Connect AWS workloads securely to core banking systems and external partners.
  • Present and defend a design, including cost and the compliance touchpoints that need specialist review.

Prerequisites

How the pieces fit

The 15-section architecture template

Every architecture in this module covers: (1) business requirements, (2) functional requirements, (3) non-functional requirements, (4) architecture diagram, (5) AWS service selection with justification and rejected alternatives, (6) network architecture, (7) security architecture, (8) API design, (9) database design, (10) data flow, (11) failure scenarios, (12) scalability, (13) cost, (14) implementation steps, and (15) compliance touchpoints needing review.

Lessons

  1. How to read and write a banking architecture — the 15-section template, design authorities and architecture decision records.
  2. Digital banking platform — channels, BFF APIs, identity, session security and the overall reference architecture.
  3. Customer onboarding — KYC/AML provider integration, secure document upload, identity verification and consent capture.
  4. Account management — account lifecycle APIs, data ownership, and customer-sees-only-their-own-accounts authorisation.
  5. Payment processing — initiation, validation, sanctions-screening hook, settlement, idempotency and ISO 20022 messages.
  6. Lab: payment saga failure scenarios — design studies 14.3 and 14.4 with a unit-tested C# saga and an optional Step Functions deployment.
  7. Real-time transaction processing — streaming authorisations, latency budgets, back-pressure and ordering.
  8. Fraud-detection integration — rules on Kinesis and Lambda, ML scoring with SageMaker or Bedrock, and human review queues.
  9. Customer notification systems — event-driven fan-out, preferences, retries and duplicate suppression.
  10. Reporting and regulatory reporting architecture — governed data lake, report lineage, sign-off and retention.
  11. Reconciliation architecture — matching internal ledgers with partner and SFTP files, breaks handling and audit trail.
  12. Secure external integrations — partner APIs, egress control, mTLS, webhooks and third-party risk.
  13. Core banking integration — hybrid connectivity, private APIs, MQ-style messaging with Amazon MQ and mainframe coexistence.
  14. API orchestration — Step Functions vs in-service orchestration, timeouts and compensation.
  15. Event-driven banking architecture — domain events, EventBridge buses per domain, schema governance.
  16. Audit and compliance architecture — CloudTrail, log archive, Object Lock, evidence collection and who-accessed-what queries.
  17. Disaster-recovery architecture — mapping important business services to DR strategies and impact tolerances.
  18. Module 14 cheat sheet — one-page revision summary.

Labs and module project

  • Design study 14.1 (🟢) — Complete the 15-section template for the payments architecture above and peer-review it with the AI assistant. Start from How to read and write a banking architecture and Payment processing. Coming: no separate walkthrough page yet; do it as an assignment.
  • Design study 14.2 (🟢) — Walk through the ISO 20022 messaging guidance from aws-solutions-library-samples/guidance-for-iso20022-messaging-workflows-on-aws and map it to your payment design. Coming: no separate walkthrough page yet; do it as an assignment.
  • Lab 14.3 (🟢, optional 🟡 deployment) and design study 14.4 (🟢) — Lab: payment saga failure scenarios. The code is in the repository folder labs/m14-payment-saga-failures/: 22 unit tests that run without an AWS account, plus an optional CDK deployment of a minimal saga on Step Functions Express and DynamoDB that you destroy afterwards. The failure workshop covers a sanctions provider outage, a duplicate client retry, a scheme timeout and a partial settlement.

Module project — architecture portfolio. Produce three complete architectures (payments, onboarding, reconciliation and reporting) using the template, each with a security-boundary diagram, cost estimate and compliance-touchpoint list. These designs underpin Capstone 1 (Secure Digital Banking REST API), Capstone 2 (Event-Driven Transaction Processing) and Capstone 3 (Banking Data Engineering Pipeline).

Cost profile

PartLabelFree alternative
All architecture studies and design assignments🟢 freeDiagrams, ADRs and documents
Payment saga proof of concept (Step Functions Express, DynamoDB; optional part of the lab)🟡 creditsUsually within free allowances for light traffic; destroy afterwards
Streaming fraud rules (Kinesis Data Streams)🟡 creditsShards bill hourly; run for under an hour, or simulate the stream locally
SageMaker/Bedrock fraud scoring🔴 paidDesign-only, or score with a local rules engine in C#
Direct Connect, Amazon MQ, multi-Region DR🔴 paidDesign-only exercise

Certification alignment

ExamDomains (weight)What this module covers
SAP-C03Design · Security & Governance · Cost (weights ⚠ VERIFY)Complex multi-service designs and trade-offs
SAA-C03Secure (30%) · Resilient (26%) · High-Performing (24%) · Cost-Optimized (20%)Service selection and justification
SCS-C03Infrastructure (18%) · Data Protection (18%)Security architecture of banking workloads

The domain labels are short forms; see each exam page for the official wording.

Service availability notes

Audit tooling — verified 2026-11-17

AWS CloudTrail Lake has been closed to new customers since 2026-05-31, and AWS Audit Manager stopped onboarding new accounts from 2026-04-30. The audit lesson uses CloudTrail delivering to S3 with Athena, and your own evidence catalogue, instead.

Fraud tooling — verified 2026-11-17

Amazon Fraud Detector is in maintenance (announced 2025-10-07) and has been closed to new customers since 2025-11-07. This module builds fraud signals with rules on Kinesis and Lambda, and optional ML scoring through Amazon SageMaker AI or Amazon Bedrock instead.

Key AWS services

Amazon API Gateway · Amazon CloudFront · AWS WAF · AWS Shield · Amazon ECS on Fargate · AWS Lambda · AWS Step Functions · Amazon EventBridge · Amazon SQS · Amazon SNS · Amazon DynamoDB · Amazon Aurora · Amazon Kinesis Data Streams · Amazon SageMaker AI · Amazon Bedrock · Amazon S3 · AWS Glue · Amazon Athena · AWS Transfer Family · Amazon MQ · AWS KMS · AWS CloudTrail · Amazon Cognito · AWS Site-to-Site VPN · AWS Direct Connect

Independent learning resource — not affiliated with, endorsed or sponsored by Amazon Web Services. AWS and related marks are trademarks of Amazon.com, Inc. or its affiliates. All banking data in examples is fictional. Sample architectures are for learning and are not automatically compliant with GDPR, PCI DSS or any banking regulation.

On this page