Procurement Platform

A full-stack procurement intake and approvals application with request management, configurable approval workflows, audit trail, and reporting dashboard.

Architecture

procurement_platform/
├── backend/          Python + FastAPI + Strawberry GraphQL + SQLAlchemy
├── frontend/         React + TypeScript + Apollo Client + Recharts
├── e2e/              Playwright end-to-end tests
├── docker-compose.yml
└── .github/workflows/ci.yml

Backend stack: FastAPI, Strawberry GraphQL, SQLAlchemy (async), PostgreSQL, JWT auth, Alembic migrations

Frontend stack: React 18, TypeScript, Apollo Client, React Router, Recharts

Quick Start

Prerequisites

• Docker & Docker Compose
• Python 3.12+
• Node.js 20+

One-command setup

bash setup.sh

Then start the services:

# Terminal 1 - Backend
cd backend && uvicorn app.main:app --reload --port 8000

# Terminal 2 - Frontend
cd frontend && npm run dev

Open http://localhost:3000

Docker-only setup

docker compose up --build

Open http://localhost:3000

Demo Accounts

Email	Password	Role
admin@acme.com	admin123	Admin
approver@acme.com	approver123	Approver
approver2@acme.com	approver123	Approver (VP)
requester@acme.com	requester123	Requester

Features

Request Management

• Create, edit, and submit purchase requests
• Track requests through their lifecycle (Draft → Submitted → Pending Approval → Approved/Rejected)
• Filter and search requests by status, category

Approval Workflow

• Rules-based approval routing with configurable policies
• Amount thresholds, category-based routing
• Multi-step approval chains
• Approver inbox with approve/reject actions and comments
• Auto-approve when no policies match

Audit Trail

• Complete audit log for all state changes
• Per-request audit timeline
• Tracks who did what and when

Reporting Dashboard

• Cycle time per status step
• Requests by category (pie chart)
• Spend breakdown by category
• Approval bottleneck view (pending items per approver)

Admin

• Create, edit, and delete approval policies
• Configure rules (amount thresholds, category matching)
• Assign approvers and set priority

Security

• JWT-based authentication
• Role-based access control (RBAC)
• Multi-tenant organization separation
• Protected API endpoints

Data Model

Organization
├── User (role: admin | approver | requester)
├── ApprovalPolicy
│   └── PolicyRule (field, operator, value)
└── PurchaseRequest
    ├── ApprovalStep
    │   └── ApprovalDecision
    └── AuditLog

GraphQL API

The API is available at /graphql with an interactive playground.

Key Queries

• purchaseRequests - Paginated list with filters
• purchaseRequest(id) - Single request with approval steps
• approvalInbox - Pending approvals for current user
• approvalPolicies - All org policies
• auditLogs - Paginated audit trail
• cycleTimeReport, categoryReport, bottleneckReport - Analytics

Key Mutations

• login - Authenticate and get JWT token
• createPurchaseRequest / updatePurchaseRequest - CRUD
• submitRequest - Submit for approval (triggers routing)
• approveStep / rejectStep - Approve or reject
• createApprovalPolicy / updateApprovalPolicy - Admin policy management

Testing

Backend unit tests

cd backend && python -m pytest tests/ -v

Frontend component tests

cd frontend && npm test

End-to-end tests

cd e2e && npm install && npx playwright install chromium
npx playwright test

The E2E test covers the full flow: login → create request → submit → approve → verify status and audit log.

CI/CD

GitHub Actions workflow (.github/workflows/ci.yml) runs:

1. Backend tests (pytest)
2. Frontend build and tests (vitest)
3. End-to-end tests (Playwright) against a live Postgres instance

Screens

1. Login - Authentication with demo credentials
2. Request List - Filterable table of all purchase requests
3. Create Request - Form to create and optionally submit a new request
4. Request Detail - Full details, approval timeline, and audit trail
5. Approver Inbox - Pending approvals with approve/reject actions
6. Policy Editor - Admin-only CRUD for approval policies with rule builder
7. Reports - Dashboard with cycle time, category breakdown, and bottleneck analysis