./osman-yigit-sokel

April 13, 202608/2025 – 02/2026

Real Estate CRM Platform — End-to-End System Design & Delivery

Designed and built a production-ready real estate CRM system to manage listings, clients, and sales workflows, owning system architecture, database design, and full-stack implementation in a startup environment.

#Backend#System Design#Crm
Client
Jokersoft
Confidentiality
Shared as an anonymized engagement summary.
Engagement
Contract
Role
Full Stack Developer
Scope
Real estate CRM system for managing listings, clients, and sales workflows · Role-based authentication and access control system · Admin dashboards for operational tracking and data visibility · Backend API layer for listings, users, and transaction management
Impact
Enabled centralized management of listings, clients, and sales workflows within a single system · Improved operational visibility through dashboard-based tracking · Established a scalable foundation for future feature expansion · Reduced manual tracking by structuring client and transaction data

Overview

Jokersoft CRM is a production-oriented real estate management system designed to handle listings, client relationships, and sales workflows within a single platform.

The system was built in a startup environment where requirements evolved quickly, requiring flexible architecture decisions and end-to-end ownership across both backend and frontend layers.

Problem Context

The core challenge was managing fragmented workflows:

  • Property listings were not centrally structured
  • Client interactions and sales processes lacked traceability
  • No unified system existed to connect listings, clients, and deals

Additionally, the system needed to support:

  • Role-based access for different user types
  • Scalable data handling as listings and clients grow
  • A clean interface for non-technical users

Solution

I designed and implemented a full-stack CRM system that centralizes real estate operations into a single platform.

Key components:

  • Backend APIs for managing listings, users, and sales workflows
  • Role-based authentication and authorization system
  • Relational database design for structured data management
  • Dashboard interfaces for operational visibility

The system enables users to:

  • Manage property listings and associated metadata
  • Track client interactions and deal progress
  • Operate within role-based permissions

Architecture Approach

The system follows a modular full-stack architecture:

  • Frontend: Next.js for building responsive UI and dashboards
  • Backend: FastAPI for API design and business logic
  • Database: PostgreSQL with structured relational schema
  • Infrastructure: Dockerized services for consistent deployment

Key Decisions

  • FastAPI over Node.js backends
    Chosen for rapid API development, strong typing, and better alignment with data-driven workflows.

  • Relational database (PostgreSQL)
    Required due to structured relationships between listings, clients, and transactions.

  • Role-based access system
    Implemented early to avoid permission-related complexity later in the system.

  • Dockerized environment
    Ensured consistent development and deployment across environments.

Constraints

  • Startup environment with evolving requirements
  • Need to balance speed of delivery with system stability
  • Limited initial system structure, requiring iterative design

What I Learned

  • Designing systems under uncertainty requires flexible architecture
  • Early decisions in data modeling heavily impact future scalability
  • Role-based systems should be planned early, not added later

Related work