Inspiration

After a year of graduate studies at Stony Brook, participating in community center activities and campus events, we noticed how much food remained afterward. Traveling on public buses also brought us closer to the everyday challenges people face, including accessing food.

Seeing surplus food alongside food insecurity led us to ask: If the food exists, what is missing between the kitchen and the person who needs it?

Our answer was coordination and accessibility. Samrit is our attempt to close that gap.

What We Built

Samrit connects food donors, distribution centers, NGO coordinators and drivers, while putting people seeking meals first.

We started with a simple question: Does the infrastructure already exist? Often, it does. Rather than replace community organizations, Samrit helps coordinate their existing people, vehicles and facilities through an adaptable workflow.

People can find venues, filter meals by allergens and preferences, get directions, and leave private feedback. An optional voice assistant makes these interactions available through conversation.

How We Built It

We built the platform with React, TypeScript, NestJS and PostgreSQL. Gemini supports menu and invoice extraction, preference interpretation and review analysis. ElevenLabs provides speech recognition and spoken responses.

QR codes, photographs and append-only handoff records help document food movement. Human confirmation remains central to food information and operational decisions.

Challenges

Our biggest challenge was making the interface simple and fast for someone seeking food while supporting detailed logistics behind it.

We also worked through voice interaction sequences, mixed food-handling requirements, accessible directions and feedback privacy. Simplifying visitor check-in from two QR scans to one reduced friction.

What We Learned

Accessibility means reducing effort, not simply adding features. Effective coordination builds on existing community infrastructure. AI can assist interpretation and operations, but accountability still requires clear records, honest limitations and human judgment.