A support workflow that is clear from start to finish
The application comes from a need close to my technical support experience: recording issues, preserving problem context and letting users check the status of a request without relying on scattered messages.
Users and scope
Users can sign in, view their tickets, create an issue and review its details and history. The demo data includes an administrator profile, but the interface does not yet implement an admin dashboard with distinct permissions.
A working demo before connecting external services
Demo authentication
Login, registration and logout to explore the product without configuring external infrastructure.
Tickets
Create, list and view issues with category, priority, status and description.
History
A ticket's detail view can display comments and events stored in the demo model.
Ready to evolve
The repository includes a schema and an optional connection for migrating persistence to PostgreSQL/Supabase.
Local persistence makes the first walkthrough reproducible
The default mode uses a JSON file and local data-access functions. API Routes isolate the frontend from storage and allow the adapter to be replaced without redesigning every screen.
Next.js App Router login / dashboard / ticket detail │ ├── React + TypeScript + Tailwind + Radix UI │ API Routes auth / tickets / comments │ ├── lib/mock-db.ts → data/mock-data.json │ └── lib/db.ts → optional PostgreSQL Deployment: Vercel demo · JSON persistence is not guaranteed in serverless environments
Technical decisions and explicit limitations
- JSON mode makes the product easier to review and the demo easier to run without external service credentials.
- The documentation separates demo mode from the PostgreSQL alternative, distinguishing configuration from a completed migration.
- The UI focuses on the core support workflow rather than simulating a business system whose permissions are not yet implemented.
- The current session uses localStorage, which is sufficient for a demonstration but requires secure cookies or an identity provider before production.
My contribution
I designed and built the application workflow, screens and API Routes, and documented the path to PostgreSQL, including the differences still to be resolved.
AI assistance with output review
This project follows my AI-assisted development workflow: breaking down requirements, exploring components, implementation, debugging and documentation. I do not present it as automatically generated software or claim that the repository includes versioned Copilot instructions that are not published there.
Fast iteration
AI helps compare alternatives and shorten the gap between a product idea and a navigable workflow.
Human validation
Routes, states, messages and limitations are checked at runtime; security decisions are not delegated to the model.
What it demonstrates today and what it does not
- The public demo supports login, dashboard navigation, ticket creation and ticket review.
- The README describes the API routes and the structure required for JSON and PostgreSQL modes.
- The repository should not yet be presented as a production system with complete server-side authorization.
- The published GitHub Actions workflow is manual and intended to fix file tracking; it is not evidence of a complete CI/CD pipeline.
- JSON persistence is not reliable for permanent changes in a serverless deployment.
Turning the demo into an operational application
The next step is to migrate the routes to PostgreSQL, secure sessions and permissions on the server, implement the actual admin dashboard, and add UI-based comments, filters, technician assignment and integration tests.