AgriLoop

Inspiration

AgriLoop was inspired by a simple problem in farming communities: sharing equipment sounds efficient, but in practice it often depends on personal trust, phone calls, and unclear handoff expectations.
We wanted to help farmers collaborate without needing a high “trust threshold” before they can safely share expensive tools.

What it does

AgriLoop helps farmers coordinate shared equipment by providing:

user accounts and authenticated access
equipment and group management
deterministic scheduling for rentals
handoff confirmation (return + receipt)
issue detection when two sides report different handoff conditions
AI-assisted risk/demo summaries
reset/seed flows for quick demos

At a high level, we aim to reduce coordination friction and make shared usage more reliable.

How we built it

We built AgriLoop as a full-stack MVP:

Backend: FastAPI + MongoDB, with JWT auth and REST endpoints
Frontend: Vue 3 + Vite
Infra: Docker Compose for local orchestration (API, MongoDB, Mongo Express, frontend)
Frontend runtime in containers: Bun for install/run workflow

We focused on deterministic, transparent flows first (schedule + handoff lifecycle), then layered in AI summaries as assistive features.

A core design idea was to make trust more measurable through event confirmation. Conceptually:

[ \text{Operational Trust} \propto \frac{\text{Verified Handoffs}}{\text{Total Handoffs}} ]

The product doesn’t replace human trust, but it lowers the amount required to get started.

Challenges we ran into

balancing speed (hackathon pace) with clean architecture
handling edge cases in handoff disagreement workflows
keeping API/UX simple while still reflecting real-world logistics
environment and container setup consistency across services
integrating AI as a helpful layer without making core flows dependent on it

Accomplishments that we're proud of

shipped an end-to-end working MVP across backend, database, and frontend
implemented a complete handoff/issue detection flow, not just basic CRUD
added deterministic schedule generation for predictability
made the stack runnable with Docker Compose for quick onboarding/demoing
kept the product focused on a concrete social/operational problem in agriculture

What we learned

trust is a product surface, not just a social assumption
deterministic systems reduce disputes because they create a shared reference point
clear data contracts between frontend/backend speed up iteration dramatically
dev velocity improves when infra and local setup are treated as first-class work
AI is most useful here as decision support, not as a replacement for core logic