BordLanka – Property
Marketplace
A comprehensive property discovery platform where homeowners showcase real estate listings, and buyers filter properties with high-performance MongoDB query boundaries.
Project Overview
BordLanka simplifies real estate listings by creating a centralized interface where property owners upload details, and searchers query listings. It reduces manual interactions by offering integrated seller contact sheets and saved lists.
The Problem Statement
Real estate web directories are frequently clogged with slow filters. Relational databases face bottlenecks when performing complex text searches across multiple tables (locations, attributes, pricing ranges) simultaneously.
The Solution Engineered
Designed a MongoDB schema optimized for flat read pathways. Added MongoDB compound text indexes on title and location fields, enabling swift matching. Implemented server-side limit parameters to fetch data in paginated steps, ensuring stable performance.
System Architecture & Orchestration
Click a Layer to Trace Data Flow
Details Panel
Responsive web client featuring high-fidelity real estate grids, custom listings creators, and dashboard interfaces for property hosts.
Key Features Demonstrated
- Dynamic client routes for property descriptions
- Instant feedback client-side search logic
- Responsive glassmorphism filters optimized for mobile
- Cloudinary asset upload scripts with image optimization
Property Listing Sandbox
Test how Mongoose query filters structure dynamic listings by searching and sorting listings on type and bedrooms.
Modern Penthouse Suite
Sunset Beachfront Villa
Cozy Studio Loft
Suburban Family Residence
Urban Executive Flat
Engineering Solutions
MongoDB Text Searching & Indexing
Problem: Basic MongoDB wildcard regex searches are expensive ($regex operations perform full table scans) and caused latency spikes when searching across thousands of properties.
Solution: Configured compound text indexes covering the `title` and `location` fields. Replaced wildcard matches with MongoDB `$text` query constraints, utilizing search weights. This reduced query execution time from 400ms to less than 15ms.
Secure Image Upload Workflows
Problem: Processing high-resolution real estate photographs directly on Node backend servers consumed memory buffers, threatening service stability.
Solution: Refactored the architecture to execute direct-to-cloud uploads. The React client requests a pre-signed cryptographic upload signature from the API server, then ships raw files directly to Cloudinary buckets, bypassing backend processing entirely.