Skip to main content

DevOps, CI/CD, and Testing

Bike4Mind’s infrastructure and development workflows are optimized for speed, reliability, and safety. The platform leverages modern serverless tooling and automated deployment pipelines, combined with rigorous testing practices and local development parity.

Infrastructure Overview​

  • Cloud Provider: AWS, using Serverless-first services (Lambda, API Gateway, S3, CloudWatch, DynamoDB, RDS where applicable)

  • Infrastructure as Code: Managed via SST (Serverless Stack)

    • Declarative infrastructure provisioning
    • Secrets management
    • Environment-specific configuration

  • Microservice-Aware Design: All deployments are modular, version-controlled, and scoped per environment (dev, staging, production)

CI/CD Pipeline​

  • Pipeline Tooling: All continuous integration and deployment is handled via SEED

    • Git-based CI/CD with preview deployments for all branches
    • One-click production rollbacks
    • Environment promotion via Git tag or branch protection rules

  • Pipeline Stages:

    1. Build: Compile TypeScript, bundle functions, lint code
    2. Test: Run unit tests, schema validations, lint checks
    3. Deploy: Deploy to dev, staging, or production environments
    4. Post-Deploy: Run database migrations (if applicable), notify Slack channels, publish version info

Local Development​

  • Developers use fully containerized or local-parity environments

  • Each developer can:

    • Run the full platform locally with access to mock or development services
    • Use .env.local configuration for secrets and environment values
    • Connect to isolated dev databases or AWS dev resources

  • Tools provided:

    • dev scripts in package.json for common workflows
    • seed dev for local serverless testing and deployment simulation
    • mongo-express, AWS CLI, or Postman for local integrations and manual testing

Testing Strategy​

  • Unit Testing:

    • Primary coverage focus is on business logic in b4m-core
    • Utilities, validators, permission logic, and API behavior are covered
    • Tests written in Jest

  • Manual QA:

    • Each sprint includes dedicated QA time
    • Exploratory testing and user scenario validation is performed on staging builds
    • Loom videos are often created to document regressions or confirm fixes

  • Automated Regression Testing (in-progress):

    • Ongoing investment in E2E testing using Playwright or Cypress
    • Future goals include visual regression testing and form interaction verification

  • Preview Deployments:

    • Every pull request generates a preview deployment via SEED
    • Allows reviewers to test the branch in a live environment
    • Facilitates rapid feedback loops between QA, design, and development

Deployment Safety​

  • Rollback Capabilities:

    • SEED enables one-click rollback to the last known good build
    • Rollback metadata is stored and logged for auditing

  • Versioning & Release Tracking:

    • Each deployment is tagged with a unique build hash and timestamp
    • Internal dashboard shows deployment history, authorship, and rollback options

  • Change Approval Process:

    • Code reviews required for all merges to main or release/* branches
    • Optional hold/release steps based on business readiness