A full-stack B2B Invoice Management Web Application built during an internship at HRC, featuring a receivables dashboard, ML-powered payment prediction, and real-time invoice operations.

In B2B commerce, seller businesses issue invoices to buyer businesses operating on credit. Tracking whether invoices are paid on time — or predicting when they will be — is a critical accounts receivable challenge. This application solves that by providing a unified dashboard to manage, visualize, and predict invoice payment timelines.

• MySQL — Relational database for invoice data
• Java + JDBC + Servlets — API layer and database connectivity
• Tomcat 10 — Servlet server
• Python 3 + Flask — Server for the ML prediction model

• React 18 — Component-based UI
• Redux Toolkit + Redux Thunk — State management
• Axios — API communication
• Tailwind CSS — Styling
• Highcharts — Data visualization (graphs)
• Material UI — UI component library

• Models evaluated: Linear Regression, SVM, Decision Tree, Random Forest, AdaBoost, XGBoost
• Techniques: Data preprocessing, EDA, feature engineering, hyperparameter tuning (Grid Search)
• Metric: RMSE, R², MSE for regression evaluation

Web Application

• Responsive Receivables Dashboard with grid and graph data visualization
• Search, Add, Edit, and Delete invoice records
• Full-stack integration: ReactJS ↔ Java Servlets ↔ MySQL

Machine Learning

• View invoice data across multiple buyers
• Perform data preprocessing and exploratory data analysis
• Predict the expected payment date for each invoice

Machine Learning — End-to-end pipeline from raw data import through preprocessing, multi-model evaluation, and hyperparameter tuning to final model selection.

Java — Core Java fundamentals, OOP, SQL, JDBC for database connectivity, Servlets for request handling, Java EE for web development.

React JS — Component architecture, JSX, props, state, hooks (useState, useEffect, useContext, useReducer), form handling, event management, and integration with Material UI, Highcharts, and Axios + Flask backend.