On-Demand Service Platform

Overview

A Django 5.2 + DRF platform that manages on-demand service orders through an explicit 9-state lifecycle with a dedicated cancellation path. Every operational event maps declaratively to its financial record, deletes are soft and recoverable, and public order IDs are cryptographically unguessable. A React 19 + Ant Design frontend consumes the OpenAPI-documented DRF API.

What I Did

• →Designed a 9-state order state machine with explicit transitions and a dedicated cancellation path
• →Implemented a SoftDeleteQuerySet so every domain delete is recoverable and auditable
• →Tuned composite database indexes around the platform's real order-lookup patterns
• →Generated public order IDs with secrets.choice for cryptographically unguessable references
• →Built PDF document generation with WeasyPrint and ReportLab
• →Mapped operational events to financial records through a declarative OPS_TO_FINANCIAL_MAP
• →Set up a singleton admin site plus a custom SuperAdminSite for elevated operations
• →Documented the API with drf-spectacular (OpenAPI) for frontend and integration consumers
• →Built the React 19 + Ant Design frontend on the DRF API

Challenges

• •Enforcing that orders only move through valid state transitions across 9 states
• •Keeping financial records consistent with operations via a single declarative mapping
• •Making soft deletes airtight so filtered-out rows never leak into queries or reports
• •Issuing public order references that are short, shareable, and impossible to enumerate

Results

• ✓Order lifecycle is fully explicit: no order can reach an invalid state
• ✓Finance and operations stay in sync through one declarative map instead of scattered signals
• ✓Soft deletes plus admin separation give full recoverability and a clear elevation boundary
• ✓OpenAPI documentation made the API self-serve for the frontend team