Smart Agro AI

The problem#

An agriculture decision-support prototype needs to bring inputs, weather context and crop recommendations into one understandable workflow.

What I built#

I built the application outside its ML component, including the React frontend and FastAPI backend. The MVP connects input forms, analysis results and weather context, while the team’s model supplies crop recommendations.

My contribution#

• Built the frontend and backend parts of the application.

• Implemented the interface for inputs and analysis results.

• Worked on the API and data flow that support the dashboard.

• Contributed to GREEN OPS as part of the hackathon team.

Behind the decisions#

Show the recommendation in context#

A dashboard connects the inputs and weather context with the output, helping someone understand the workflow rather than only seeing a prediction.

Be clear about demo behavior#

The backend distinguishes model inference from a simulation fallback. The application remains an MVP, with model validation and dataset provenance still limited.

What I learned#

Building around a team’s ML component helped me understand the boundaries between a model, an API and the interface that presents its output.

What comes next#

• Improve the end-to-end analysis experience and its error states.

• Continue strengthening the interface and backend.

• Work with the ML contributor on dataset provenance and validation before considering real-world use.