Grō | Devpost

Landing Page
Login Page
Upload File Page
Soil Analysis Page
Final Result Page

Inspiration

One of our member came from Malang, a city with quite a lot of farmers. Growing up, that member became quite acquainted with farmers and the challenges that they are facing. For farmer that owns small fields, the margin that they earned is so small compared to the risk that they take. For them, deciding on what to plant for the season means a lot. Hence why, we decided to help farmers decide on what they plant based on factors that affected the plant growth.

What it does

Gro is a mobile-first precision agriculture tool that provides personalized crop recommendations. Using just a photo of the soil and local weather data, Gro analyzes soil texture, fertility, and climate conditions to suggest the best crop to plant, the ideal time to do it, and how to optimize harvest outcomes. Farmers no longer need agronomists or costly sensors—Gro turns every phone into a smart farming assistant.

How we built it

We started by developing the core features of the system—an AI-powered crop recommendation engine and a soil photo analysis module. Once these key components were functional, we proceeded to build supporting features such as user authentication, session management, and login/logout capabilities. Finally, we implemented a recommendation history feature to allow users to track and review past planting suggestions. Our designer crafted the UI/UX screens iteratively, which were then implemented into the application frontend. Once the foundational features were in place, we integrated the backend and frontend components to ensure a seamless user experience. The application was deployed to a dedicated server, and we set up a CI/CD pipeline to streamline development, and deployment processes efficiently.

Challenges we ran into

The app behaves differently between the local environment and production (server) environment. Hence why it was quite a challenge to make the app works on server with limited time.

Accomplishments that we're proud of

We learned how to manage the context within a request into LLMs and external API. This is something we have never tried before, but now we have a reason to and we are proud of how fast we learned that.

What we learned

We learned how to discuss, brainstorm, and develop within a short amount of time. We also learned a lot on what can be done better when developing a product within a short amount of time.