InmoPipeline
Provides an end-to-end pipeline for real estate market analysis and predictive modeling, encompassing data collection, transformation, visualization, and machine learning.
About
InmoPipeline is a comprehensive end-to-end data engineering and data science project designed for the Colombian real estate market. It integrates automated web scraping, robust ETL processes, and efficient data warehousing with DuckDB. The project includes extensive data analysis, interactive dashboards built with Streamlit, and a powerful Random Forest machine learning model for property price prediction. The solution is deployed as a REST API using FastAPI and features intelligent automation workflows via n8n, demonstrating a full data product lifecycle from ingestion to intelligent querying.
Key Features
- Automated Web Scraping and ETL Pipeline for real estate data
- Interactive Dashboards for market analysis and KPIs using Streamlit
- RESTful API for price predictions and data queries deployed on Render
- Predictive Machine Learning Model (Random Forest) for property prices with 70.07% R²
- 0 GitHub stars
- Intelligent Automation Workflows with n8n and Google Gemini LLM for AI-powered real estate inquiries
Use Cases
- Automating real estate market analysis and insights generation
- Predicting property prices based on various features and market trends
- Enabling automated, intelligent queries about real estate properties and market data
