A family celebration in June got Vihaan Das thinking about food waste. His brother was having a party, and his mother was unsure how much food to order and prepare the guests, and the problem caught Vihaan's attention. He wanted to work on a tool to help with the calculation, and his classmate Krishiv Kadam joined him in developing SmartServe, an application that runs on a computer and estimates food quantities based on the number and preferences of guests. By August, the Grade VIII students at JBCN had developed an initial version.
Users enter how many guests are vegetarian or non-vegetarian and classify them by age: adults, teenagers, children, and babies. They then select the dishes on the menu, including main courses, side dishes, desserts and snacks, and choose drinks. A slider lets them set a calorie limit. The app calculates the quantities required for the selected menu and retrieves nutritional information, including calorie counts, through the FatSecret API, an external database of food and nutrition information. It also uses stored estimates when information on certain common foods is unavailable.
Vihaan wrote the code in Python and built the application's interface using Tkinter, a Python tool for creating desktop applications. Its calculations rely on formulas he devised to assign different food quantities to each age group. The starting estimate for rotis, for instance, is 3.5 per adult, 2.5 per teenager and 1.5 per child. The combined figure is then multiplied by 0.95, a small adjustment to the estimate. Other dishes have their own formulas.
Krishiv's contribution centred on understanding the problem the application was intended to address. He researched food wastage and dietary preferences and prepared questionnaires to understand how people perceive the problem and what steps they take to prevent it. The two students also had to decide how much food their calculations should recommend. "We wanted to ensure that while no food went to waste, no one went hungry either. At the end we decided to settle on a middle point," Krishiv explains.
As one would expect, that balance remains difficult to calculate — although users can select different dishes, the current version does not adjust its estimates according to whether an event is a full meal or a gathering serving only snacks. It also does not fully account for serving sizes, event duration, individual appetites, or how many guests might choose each dish. According to Krishiv, SmartServe is currently suited to small gatherings with a limited menu. It can take two to three minutes to process entries and may produce inaccurate estimates for larger groups.
The students have tested the application with peers and teachers, using their feedback to improve the calculations. However, it has yet to be tested at an actual event, Vihaan points out. "I would test SmartServe at several gatherings and record the number of guests, the amount of food prepared, the amount eaten and the amount left over. I would weigh the leftovers and calculate the percentage of prepared food that was wasted. I would then compare gatherings that used SmartServe with similar gatherings that did not use it," he says.
The students also want to add an event-type option, allowing the application to account for the difference between a full meal and a snacks-only gathering when estimating quantities.
