At large Indian weddings, food remains one of the most visible markers of scale and spending. Caterers routinely present dozens of dishes across multiple counters, reflecting regional traditions, family status, and the expectation of abundance. In recent years, this format has become nearly standard across urban centres such as Bengaluru, Delhi, Mumbai, and Hyderabad, where destination venues and professional event planners manage gatherings that often include hundreds or even thousands of guests.
While the volume and variety of food are intended to signal hospitality, they have also created a recurring challenge for attendees who must make quick decisions in crowded spaces with limited time and physical capacity. This everyday experience became the starting point for a recent technology experiment that gained wide public attention after being shared online by a Bengaluru-based software professional.
The experiment, informally named “BuffetGPT,” was introduced by Pankaj Tanwar, a software engineer working in Bengaluru. Tanwar described the project as a personal attempt to address the dissatisfaction he often experienced while navigating wedding buffets. His description of the idea appeared on the social media platform X, where he referred to the traditional Indian wedding buffet as inefficient for decision-making.
indian wedding buffet is a scam. i always leave regretting something. so i built BuffetGPT ?
an ai agent that scans entire buffet and gives you a game plan.
it uses computer vision to detect every dish, then optimizes what to eat, what to skip, and how much based on actual… pic.twitter.com/yLxEf17LSJ
— Pankaj (@the2ndfloorguy) February 6, 2026
The post quickly circulated across technology and general interest communities, drawing responses from software professionals, wedding industry workers, and members of the public familiar with large social gatherings. The attention placed the project within a broader discussion about the application of computational systems to routine social settings.
Indian wedding catering has expanded substantially over the past two decades, driven by rising disposable incomes and the professionalisation of event services. Industry estimates show that wedding-related spending in India runs into several billion dollars annually, with food and venue costs forming a large share of overall budgets. Catering companies often compete by offering larger menus, live cooking stations, and region-specific dishes intended to appeal to diverse guest lists.
While hosts and planners frame this abundance as a benefit, guests frequently face queues, overlapping meal courses, and uncertainty about portion sizes. Tanwar’s project emerged from this setting rather than from a commercial initiative or institutional research programme.
According to Tanwar’s public statements, BuffetGPT was designed to examine a buffet spread and generate a structured eating plan for an individual guest. The system, as described, relies on visual recognition to identify dishes placed across serving stations. Tanwar said the programme categorises items based on rarity, preparation effort, and how commonly they are consumed outside wedding settings. Items considered common household foods are placed in a lower priority group, while dishes prepared specifically for large events are placed higher. The intent, according to his explanation, was to guide guests toward choices they were less likely to encounter in everyday meals.
Tanwar stated that the system also takes into account physical constraints, including the amount of food a person can reasonably consume during one sitting. He referred to this aspect as an attempt to account for stomach capacity rather than appetite alone. Based on the available description, the system produces a sequence suggesting the order in which dishes should be tried and approximate portion sizes. Tanwar did not release technical documentation, datasets, or performance metrics, and no independent verification of the system’s effectiveness has been published.
A Bengaluru-based software professional has created an unusual AI tool to solve a problem familiar to many Indians—choosing the right dishes at an overloaded wedding buffet. The tool, called BuffetGPT, helps guests plan what to eat, what to skip, and how much to consume to avoid… pic.twitter.com/OhOR1sLrbc
— News9 (@News9Tweets) February 7, 2026
The project remains at an early stage. Tanwar said the system was tested once during a private wedding event involving a limited number of participants. He described the outcome as acceptable but did not provide quantitative results or user surveys. There has been no public demonstration beyond images and brief explanations shared online. The project has not been commercialised, and Tanwar has not indicated any plans to sell or formally deploy the system at events.
Public reaction to the idea was immediate and varied. Many online responses focused on the novelty of applying computational analysis to a familiar cultural setting. Users commented on the shared experience of regret after overfilling plates or missing popular dishes once queues formed. Some suggested extensions, such as linking the system to caterer ratings or guest preferences. Others questioned whether such a system could operate reliably in crowded venues where dishes change frequently, and serving stations may not follow a fixed layout.
Several responses highlighted practical constraints. Commenters noted that social dynamics at weddings, including pressure from relatives or hosts, often dictate eating patterns regardless of planning. Others pointed out that dishes may run out quickly, making any pre-set plan difficult to follow. These reactions underscored the difference between structured decision systems and the unpredictable nature of large social gatherings.
Tanwar’s online profile indicates that BuffetGPT is not his first public technology experiment. In previous posts, he described building a system that automatically switched his computer screen from entertainment content to work-related material when a supervisor approached his desk. That project was framed as a personal solution to workplace visibility rather than as a product. More recently, Tanwar attracted attention for a helmet-mounted camera system designed to record traffic violations and submit reports to authorities, a project that later drew an official response from the Bengaluru City Police, who said they were interested in understanding the concept further.
In the case of BuffetGPT, Tanwar has consistently described the work as exploratory. In his posts, he said the project was an exercise in applying technical skills to everyday problems rather than an attempt to solve a major social issue. He did not claim that the system would change how weddings are organised or how catering is planned. His comments emphasised personal curiosity and experimentation.
Now A ‘BuffetGPT’? The Food Planner That Knows What You Should Eat At Weddings!https://t.co/VmADncZjwZ
— The420.in (@The420in) February 7, 2026
The episode also drew attention from technology commentators who view such projects as examples of engineers applying analytical frameworks to social experiences. While no institutions or companies have adopted the idea, its circulation online placed it within ongoing discussions about how computational tools intersect with cultural practices. Weddings in India remain highly ritualised events shaped by tradition, family expectations, and social norms, areas that are often resistant to formal planning systems.
