Project Overview
NeighborIQ is an AI-powered intelligence platform that transforms raw real estate data into actionable neighborhood insights. It acts as a unified microservices ecosystem for the Canadian residential market, combining automated data ingestion with machine learning to provide price predictions and investment analytics. Built for extensibility and performance, it leverages a distributed architecture to deliver enterprise-grade market transparency without high operational overhead.
Features
- Unified Property Gateway: Single API entry point for multi-dimensional property search and management.
- ML Price Predictions: XGBoost-powered inference engines for property valuation and rental yield forecasting.
- Automated Data Pipeline: Scrapy-based ingestion framework for real-time Canadian market listing updates.
- Geo-Spatial Search: Elasticsearch integration for high-performance location-based and full-text property discovery.
- Enterprise Authentication: RS256 JWT tokens with refresh rotation and JWKS-aware gateway security.
- Intelligent Caching: Multi-layer Redis strategy for search results, sessions, and asynchronous task results.
- Async Task Processing: Celery-driven background workers for heavy lifting in ML and web scraping.
- Portfolio Management: User-specific watchlists and saved property tracking for investment monitoring.
Tech Stack
Backend & AI
- Framework: FastAPI (Async-first)
- Language: Python 3.11
- ML Engine: XGBoost & scikit-learn
- Database: PostgreSQL 15 + PostGIS
- Search & Cache: Elasticsearch 8.11 & Redis 7
Frontend
- Framework: Vue 3 + TypeScript
- Styling: Tailwind CSS
- State & Maps: Pinia & OpenLayers
Infrastructure
- Containerization: Docker & Docker Compose
- Reverse Proxy: Nginx
- CI/CD: GitHub Actions with Trivy security scanning