How sanjith-3/agrisat-ai works
Project type and purpose
• CropSense is a lightweight, single-page, full-stack–style web application built in Python using Streamlit.
File structure
• Top-level:
– .devcontainer/ (development container)
– stream.py (the Streamlit app entrypoint)
– requirements.txt (Python dependencies)
– README.md (project overview, instructions)
• No separate “frontend” or “backend” folders—the Streamlit server doubles as UI and application logic host.
• No persistent database or microservices—everything runs in one process.
README for additional insights
• Describes these logical steps:
1) Upload a zip of TIFF bands
2) Stack bands into arrays
3) Compute 13 indices via Rasterio/EarthPy
5) Send index data to Groq AI and display returned advice
6) Delete temp files
• Tech stack: Python 3.8+, Streamlit, Rasterio, EarthPy, Groq API
Diagramming recommendations
– UI (blue), Application logic (green), Data processing (orange), External services (purple), Storage (gray).
Details to capture
• Which libraries handle each step.
• Single-process nature but logically separate concerns—this makes the monolith understandable at a glance.