Platform features

See Licence tiers and platform configuration for capability configuration.

Identity and security

  • Service-to-service authentication (JWT / Cognito service pool)
  • User authentication (Cognito JWT, email/password, first-party UI)
  • User authentication (Google OAuth2)
  • User authentication (LinkedIn OAuth2 + account linking)
  • Registration / token refresh / password reset
  • Service-level authorization (named caller allowlists)
  • Zero-trust request model (stateless JWT, no server sessions)
  • Secrets management (Secrets Manager / SSM Parameter Store; no long-lived AWS keys in CI)
  • GitHub Actions OIDC to AWS deploy roles
  • Edge protection (CloudFront + WAF: host allowlist, geo, rate rules)
  • Origin protection (ALB locked to CloudFront + verify secret)
  • TLS-only viewers and HSTS at the public edge
  • Private ECS networking / VPC per environment
  • Runtime isolation (ECS Fargate, private subnets, non-root container user)
  • Encryption at rest with AWS-managed keys (AMK)
  • DNSSEC on Route53 hosted zones
  • CSP / XSS posture for browser token storage
  • [Growth + Enterprise licence tier] Customer-managed KMS (platform-provisioned CMKs per encryption domain)
  • [Enterprise-only licence tier] Bring-your-own-key (BYOK) ARNs per encryption domain
  • [Enterprise-only licence tier] Optional internal ALB TLS (east-west HTTPS on the internal listener)

Traffic control and resilience

  • Application throttling / rate limiting (Redis-backed, distributed)
  • Edge throttling (CloudFront WAF)
  • Distributed caching (ElastiCache Redis; cache-aside)
  • Circuit breakers / fault tolerance on persistence paths
  • Dependency-aware health / readiness (Postgres, DynamoDB, S3, Cognito, SES, Redis)
  • [Enterprise-only licence tier] HA endpoint AZ mode (ALB and VPC interface endpoints spanning AZs)

Observability

  • Prometheus metrics (HTTP, JVM, domain ops, throttle, DB, circuit breakers)
  • Grafana dashboards (local/lab)
  • Distributed tracing instrumentation (OpenTelemetry)
  • Structured / correlated logging (MDC, X-Correlation-Id)
  • PII / sensitive-data log masking
  • CloudWatch + SNS infrastructure monitoring / alerting
  • [Growth + Enterprise licence tier] Optional managed observability (AMP remote write + AMG + X-Ray OTLP export)

Domain services

  • auth-service (identity lifecycle)
  • actor-service (user profiles + caching)
  • audit-service (central ingest + Postgres persistence)
  • Declarative cross-cutting audit emission
  • notification-service (async email via SES, templates, retries, unsubscribe tokens)
  • document-service (upload/storage + parsing across SQL / DynamoDB / S3)
  • template-service scaffolding enables RAD
  • BFF orchestration (actor / admin persona APIs)
  • Stateless reference UI (no client-side state; HATEOAS-aware; static CloudFront/S3 path)

Data and messaging

  • PostgreSQL (profiles, audit, notifications, relational workloads)
  • DynamoDB (high-throughput / document-oriented shapes)
  • S3 object storage
  • SES email delivery
  • Guaranteed audit message delivery
  • EventBridge audit fan-out with SQS ingest (SIEM rules attach as operator extension)

Platform delivery and infra (AWS)

  • AWS CDK IaC (VPC, DNS, certs, WAF, ALB, ECS Fargate, ECR, datastores)
  • ECS/Fargate deploy model + autoscaling
  • GraalVM native image builds (with JVM fast-jar escape hatch)
  • GitHub Actions CI/CD spine (hygiene → build/test → static analysis → coverage → version → ECR → env deploy)
  • Multi-env deploy workflows (int / stage / prod)
  • Non-prod infra hibernate (FinOps)
  • Optional NAT Gateway for private-subnet egress
  • Optional self-hosted GHA runners (native builds)
  • k6 performance test framework

Governance and compliance foundations

  • Application audit trails
  • Compliance control mapping (SOC 2 / GDPR / HIPAA-aligned objectives; operator-owned certification)
  • Compliance roadmap / shared-responsibility model
  • Least-privilege IAM posture documentation; exceptions documented (nag)
  • Edge access logging (CloudFront and static-content stores)
  • VPC reject flow logs (PROD; one-year retention)
  • [Enterprise-only licence tier] Regional CloudTrail (deployment region plus global IAM/STS) with log-file integrity validation
  • [Enterprise-only licence tier] Protected evidence storage (versioning, lifecycle, S3 Object Lock COMPLIANCE, stage-aware retention)
  • [Enterprise-only licence tier] ALB access logs into a companion evidence bucket

Disable the CloudTrail, Object Lock, and ALB access-log capabilities when your organization already operates org-wide trails and log archives.

Developer experience

  • Floci-powered local development
  • Taskfile ops automation plane and cheatsheet
  • Opinionated Quarkus/Java service layout (clean architecture)
  • Maven dependency management (Quarkus/AWS BOM imports, aggregator-scoped pins, versionless leaf POMs)
  • Optimised build and test wall-clock (Maven build cache, parallel reactor, parallel Surefire/Failsafe; CI typically under 10 minutes)
  • OpenAPI (e.g. actor-bff)
  • Renovate / dependency hygiene workflows
  • Architecture guides
  • Architecture Decision Records (ADRs)
  • Published performance envelopes

Shared platform (backbone-kit)

Open-source libraries for observability, throttling, health, and small cross-cutting APIs - so product repos stay about domain, not plumbing:

ModuleWhat it brings
backbone-metrics / backbone-metrics-apiDeclarative application metrics (operations, DB, resilience, throttles)
backbone-health-awsReadiness for AWS-shaped dependencies (Postgres, DynamoDB, S3, Cognito)
backbone-http-awsSigV4-signed HTTP transport (AMP, X-Ray OTLP, other AWS endpoints)
backbone-observability-api / backbone-observability-awsAMP metrics push (Prometheus remote write) and X-Ray trace export (OTLP + SigV4)
backbone-throttleRate limiting primitives and metrics-friendly integration
backbone-securityReusable JWT building blocks (Cognito claims extractors ordered in Backbone)
backbone-loggingMethod-entry logging, correlation IDs, PII / sensitive-data masking
backbone-commonShared REST, validation, and error-handling utilities

Licence tiers and platform configuration

Your signed licence tier (foundation, growth, or enterprise) determines which platform configuration options the bootstrap procedure allows. Settings apply per environment (DEV, INT, STAGE, PROD) so integration stages can stay lean and production-like stages can take a fuller posture, keeping FinOps predictable.

ConfigurationFoundationGrowthEnterprise
vpcAzCountmax 2max 2unrestricted
endpointAzModeminimalminimalminimal | ha
natEnabledtrue | falsetrue | falsetrue | false
ecsDesiredCount1–21–3min 1, no max
internalHttpsfalsefalsetrue | false
customerManagedKmsAMKAMK, CMKAMK, CMK, BYOK
observabilityEnabledfalsetrue | falsetrue | false
governanceEvidenceEnabledfalsefalsetrue | false

Managed observability (observabilityEnabled) gates the AMP/AMG stack and application export to Prometheus remote write and X-Ray.

Infrastructure monitoring (CloudWatch dashboards and alarms to SNS) and application logs in CloudWatch Logs ship on every tier regardless.

Addition of a NAT Gateway for outbound internet access is a configurable option for all tiers (with cost-saving implications).

Encryption at rest defaults to AWS-managed keys (AMK). Growth adds Backbone-provisioned CMKs (CMK); Enterprise adds bring-your-own-key (BYOK) per domain.


Did this page help you?