At Mahaveer Marketing, I architected a full-stack quotation management platform for a garment label and price tag manufacturing business. It combines two machine learning models with tools for managing quotations, customers, and business analytics.
I engineered a dual-stage prediction pipeline using about 32,000 quotation records and 13+ input features to estimate manufacturing cost and recommend selling prices. I also developed REST API modules and structured the PostgreSQL database using Supabase.
In my Real-Time Accident Risk Prediction & Geospatial Analysis System project, I combined DBSCAN clustering and Isolation Forest to identify high-risk accident zones. I also built a Flask API and interactive Leaflet.js dashboard with weather data and click-to-predict functionality.
