Building Portfolio Projects Beyond CRUD Operations
Many portfolio projects stop at basic CRUD (Create, Read, Update, Delete) functionality. This post documents how to build projects with richer features: transaction management, authentication, event processing, integration testing, and containerization—all of which demonstrate deeper engineering skills to potential employers.
Project Requirements
A portfolio project should go beyond a simple to-do list or note-taking app. Requirements should include: multi-tenant data isolation, role-based access control, event-driven workflows, and export/import functionality. These features demonstrate the ability to build software that is secure, scalable, and maintainable.
Architecture & Data Modeling
The project uses a modular architecture with separate modules for core, auth, billing, and reporting. Data modeling includes relational tables with proper foreign keys, indexes for query performance, and migration scripts. Entity-relationship diagrams are generated using SQLModel and Alembic for versioned schema changes.
Error Handling
A custom exception hierarchy handles expected and unexpected errors:
class AppError(Exception): status_code = 400 def __init__(self, message, status_code=None, payload=None): super().__init__(message) if status_code is not None: self.status_code = status_code self.payload = payloadAll endpoints wrap return values in a consistent JSON format:
{"success": bool, "data": ..., "error": {...}}.Authentication
JWT-based authentication with refresh/access token flow. Passwords are hashed using bcrypt. Token expiration is 15 minutes for access tokens and 7 days for refresh tokens. Protected endpoints use a decorator:
@auth_required.Event Processing
Kafka is used to publish domain events:
UserRegistered,PaymentCompleted, andPasswordResetRequested. A consumer service listens and triggers side effects: sending welcome emails, updating analytics, and generating reset links.Testing
Integration tests cover the full request lifecycle: authentication, event production, and consumer processing. Testcontainers spin up a Kafka broker and PostgreSQL instance. Unit tests use pytest with parameterized inputs for error handling paths.
Deployment
The project is containerized with Docker Compose: a web service, PostgreSQL, and Kafka broker. CI/CD pipeline uses GitHub Actions to run tests on every push and deploy the Docker image to DigitalOcean App Platform on merge.
Repository Evidence
Project source code is published on GitHub with a README that includes:
- Setup instructions (Docker, environment variables).
- API documentation (Swagger/OpenAPI).
- Database schema (ER diagram).
- Run book for local development.
