IT Service Desk & Asset Inventory Platform — A full-stack, multi-tenant application built entirely with AI-assisted development to demonstrate how production-grade software can be designed, implemented, and validated using AI as a development partner.
DeskSupportMonkey is a working IT service desk where companies manage help desk tickets, IT asset inventory, users, departments, and reports. It supports multi-tenancy, role-based access, real-time notifications, PDF reports, and a modern React frontend.
This project demonstrates a structured AI development methodology. Every epic was developed through an AI agent pipeline that mirrors a professional software team's workflow.
| Role | Capabilities |
|---|---|
| Employee | View equipment, submit requests (6 types + AI classification), track requests, schedule appointments, view shipments, real-time notifications |
| Technician | Request queue, asset inventory with QR/barcodes, maintenance calendar, shipment management, availability settings |
| Admin | Dashboard with metrics/SLA alerts, user/department management, equipment profiles, procurement, maintenance templates, PDF reports, API keys |
| Super Admin | Multi-tenant company management, status control |
- 20+ bounded contexts following DDD + CQRS + Clean Architecture
- AI-powered request classification and auto-assignment (Groq/OpenAI)
- MCP server with 57+ AI-callable tools for Claude Desktop
- Real-time WebSocket notifications, domain event bus
- Procurement: POs, vendors, budgets, supply chain risk, contracts
- Compliance: NIS2/DORA/ISO 27001 dashboards, audit trail, risk register
- Change management (ITIL), security incident lifecycle, vulnerability tracking
- Background PDF reports (Celery + MinIO), bilingual UI (EN/ES)
- Magic link + password + Google/Microsoft OAuth2 authentication
Backend: Python 3.13, FastAPI, SQLAlchemy, Alembic, Celery + Redis, JWT, MinIO (S3), uv Frontend: React 19, TypeScript 5.9, Vite 7, Tailwind CSS 4, TanStack React Query, React Router 7 Infrastructure (Docker Compose): PostgreSQL 15 (:5443), Redis (:6398), Mailpit (:8028/:1028), MinIO (:9000/:9001)
Each bounded context follows DDD layers:
{bc}_bc/{subdomain}/
├── domain/ # Entities, enums, repository interface (port)
├── application/ # Commands (write) and queries (read) — CQRS
└── infrastructure/ # SQLAlchemy models, repository implementation
HTTP adapters in adapters/http/api/ orchestrate commands/queries. Dependency rule: Domain ← Application ← Infrastructure ← Adapters.
# 1. Clone and configure
git clone <repo-url> && cd desksupportmonkey
cp .env.example .env
# 2. Start infrastructure
make start-docker
# 3. Install dependencies, migrate, seed
make install && make db-upgrade && make seed
# 4. Start everything
make start| Service | URL |
|---|---|
| Frontend | http://localhost:5173 |
| API Docs | http://localhost:8000/docs |
| Mailpit | http://localhost:8028 |
| MinIO | http://localhost:9001 |
| Company | Role | |
|---|---|---|
| (global) | Super Admin | admin@desksupportmonkey.com |
| TechCorp | Admin | alice.smith@techcorp.com |
| TechCorp | Technician | bob.johnson@techcorp.com |
| TechCorp | Employee | dave.brown@techcorp.com |
make start / stop # All services
make start-backend # FastAPI only
make start-frontend # Vite dev server
make queue # Celery worker
make test # Unit tests
make test-integration # Integration tests (Docker required)
make lint # mypy + flake8
make db-upgrade / db-migrate msg="x" / db-downgrade
make seed / demo-reset # Load / reset demo data
make scan # OWASP ZAP security scan10 agents in ai_docs/agents/ form a complete SDLC — from requirements to validation:
Requirements: Write → Validate → Slice → Design → Tasks Validation: DoD Check → Architecture → Quality → Performance → Linter → Testing
Architecture standards enforced by agents are documented in ai_docs/architecture/.
Built by Juan Macias.