The imperative to tackle hunger extends beyond mere sustenance, encompassing broader societal concerns and welfare. It also involves tackling food shortages, reducing malnutrition, and bridging the gap between scarcity and surplus. However, in societies with abundant resources, the paradox of excessive food waste exacerbates hunger and related issues.

QUAD.B addresses these challenges by offering solutions to minimize food waste, optimize raw material purchases based on consumer demand, and facilitate the redistribution of surplus items to those in need. We aim to ensure optimal resource utilization, provide food at affordable rates, and contribute to the goal of zero hunger. We present a seamless web application providing features for end-users to predict food wastage and collaborate to maximize food redistribution. Along with our highly database-driven web application, we also deploy an XGBoost Regressor model to predict food wastage based on user input features.

• View Implementation

• Key-Feature Breakdown:
  
  ◦ Wastage Prediction using Machine Learning: Utilizing machine learning algorithms, our platform predicts food wastage scenarios based on historical data and user inputs. We experimented with various regression models like XGBoost, CatBoost, Decision Tree, Linear and Logistic Regression.
  
  ◦ Feed: This serves as a marketplace offering a wide range of consumable items at reduced prices. It encompasses various products, including clearance items obtained from nearby stores and freshly prepared meals from certified restaurants. In addition to this, the marketplace may feature a variety of other goods and products, providing users with diverse options at affordable rates.
  This feature has been envisioned to introduce a social-media alike view to enhance user interaction on listed items, be they freshly prepared food or stored items at discounted rates. Users can apply advanced filters to get a specific product without a need to visit every store, or restaurant to check if they are offering it or not.
  This implementation brings in an in-house shopping experience from different stores in the locality and then picks them up based on users' convenience. The applications allow users to place orders using the in-house payment gateway and also a pay-upon-pickup for ease of users.
  
  ◦ Partner profiles and Order Management: This feature allows users to interact with partners, such as stores, restaurants, and donors. Users can view detailed profiles of these partners, including information about their offerings, location, and contact details. We also enable users to place and track orders seamlessly, facilitating efficient transactions between users and partners. Overall, this functionality enhances user engagement, fosters trust between users and partners and streamlines the process of accessing goods and services through our platform.

• Technologies Used:
  ◦ ReactJS
  ◦ Express.js, Node.js
  ◦ MongoDB
  ◦ ANN, XGBoost, Catboost and other Machine Learning Techniques
  ◦ Encryption (BCrypt and RSA) for authorization