Selected work

Case study 03 / Organizational systems · Integration

Zaraamad.

A connected backend for customer operations, municipal services, pricing, and support across existing systems.

Full-stack Developer

PythonFastAPIDjangoPostgreSQLSQL ServerSQLAlchemyAlembicDocker

Overview

The context behind the code.

Zaraamad brings together organizational and municipal service workflows, including service configuration, customer operations, and revenue-related information. Its public backend repositories show both the application domain and the operational tools needed to work with distributed customer databases.

The problem

The problem

Customer-specific services, pricing, subscriptions, and support cannot be treated as unrelated records. The backend also has to coexist with established authentication and database systems while keeping access rules and changing business requirements understandable.

Architecture

Architecture

The portal uses FastAPI modules with explicit controllers, services, schemas, and data mappings over SQLAlchemy and PostgreSQL. Separate bridge clients connect existing services, while Django support tooling routes connections to city databases and a Python query runner produces reports across SQL Server instances.

SYSTEM OVERVIEW
Portal & support01
Domain services02
Integration bridges03
Customer databases04
Conceptual view · Read the case study for implementation context

The main Zaraamad backend is a private company system; the linked portal, support, and query-runner repositories are public parts of it. The diagram is a conceptual view of their responsibilities.

My contribution

My contribution

I built several of Zaraamad's municipal revenue subsystems: trade duties for union and non-union businesses, transferred-income duties, and the general "Other" subsystem, which accepts any type of municipal duty, calculates it, and issues the bill to the citizen. I also built search across the very large volume of citizens' bills in these subsystems. About three quarters of the commits in each of these subsystems are mine, from June 2023 to September 2026.

Engineering challenges

Engineering challenges

The portal verifies Django-issued JWTs while serving its own FastAPI endpoints, making compatibility at the authentication boundary explicit. Customer-specific configuration, remote bridge availability, and partial failures across multiple databases add another layer of engineering beyond a single application database.

Engineering reflections

Engineering reflections

Each kind of municipal duty has its own rules, but citizens and staff need them to behave consistently; a general subsystem for any duty type only works when those rules stay explicit and inspectable.

Technology

The tools behind the work.

  • Python
  • FastAPI
  • Django
  • PostgreSQL
  • SQL Server
  • SQLAlchemy
  • Alembic
  • Docker
Next case studyZarvand