I enjoy building scalable backend systems and AI-powered applications that solve real-world problems.

github

Aditya Dwivedi  • Peerlist

Developed an AI-powered real estate intelligence web application tailored for Indian markets. The prediction engine utilizes a weighted linear regression simulation that factors in over 15 city multipliers, 8 property types, base area metrics, and specific condition and amenity bonuses to forecast property values with a dynamically calculated confidence score. Key features include: • AI Prediction & Comparison: Allows users to generate highly customized price estimates and run side-by-side comparisons of two properties to receive a winner recommendation. • Interactive Market Dashboard: Built with Chart.js to visualize 12-month market trends, city comparison bars, and comprehensive EMI loan breakdowns. • Neighborhood Insights & Investment Scoring: Calculates demand/supply indices, rental yields, and an overall 0-100 investment attractiveness score. • Modern UI/UX: Designed a responsive, luxury dark-themed interface with smooth animations and intuitive price breakdown visualizations.

Advanced House Price Prediction

Finding the right place to live can be overwhelming, so I decided to build a solution. FlatFinder is a comprehensive, user-friendly real estate platform designed to make discovering flats and apartments across India seamless and efficient. As a Computer Science student at Dr. APJ Abdul Kalam Technical University, I love building projects that solve real-world problems. With FlatFinder, I wanted to focus on creating a robust backend architecture paired with an intuitive user experience. Key Features of the Platform: • Advanced Search & Filtering: Browse listings tailored to specific budgets, locations (Kanpur, Mumbai, Delhi, Bangalore, etc.), and required amenities. • Interactive Map View: Visualize property locations easily to understand neighborhoods better. • User Authentication: Secure login and sign-up functionality for both property seekers and verified owners. • Save & Compare: Functionality to favorite listings and compare them at your convenience. • Direct Communication: Built-in capability to directly contact owners or agents to schedule visits.

FlatFinder! 🏢

As I continue diving deeper into Data Science and predictive modeling, I recently built and deployed a robust ensemble model designed to forecast rainfall with high precision based on meteorological parameters. This project was a fantastic opportunity to integrate advanced machine learning techniques with a solid backend architecture, turning complex weather data into a fully interactive and deployed web application.