I’m a 19 y/o CSE student building Satvian — a field-level AI platform for Indian farmers. Would love feedback from people working in AgTech
Hey everyone, I'm a 19-year-old CSE student from Punjab, India, and I've been building an AgTech project called Satvian mostly on my own.
I started it because I kept seeing agricultural information scattered across different places — weather in one app, mandi prices somewhere else, crop advice somewhere else, satellite data somewhere else — while a farmer often just wants to ask:
“What should I do with my field today?”
So I tried building the whole workflow around the farm itself.
A farmer can ask questions through text or voice in English, Hindi or Punjabi, and Satvian combines things like:
- field location and area
- crop and crop stage
- soil information
- weather and rainfall
- satellite-derived crop health indicators
- historical field observations
- government mandi data
- agronomy documents from agricultural institutions
The goal isn't just to return a generic chatbot answer. For example, instead of simply saying “wheat needs irrigation during tillering,” the system can take the farmer's actual field conditions, recent rainfall, forecast, soil information and satellite indicators into account before recommending an action.
I'm also working on the market side. Satvian can compare mandi prices across districts and consider the farmer's location and travel/transport cost, so the question becomes more like “where would I actually make more money selling this crop?” rather than just showing a price table.
I've also built things around it such as:
field polygons → satellite monitoring → crop health history → AI advice → farm memory → mandi intelligence → farmer community
The app is already live on Android, and I've had users from multiple countries testing it even though the main focus right now is India.
I'm currently in the testing/validation stage, and I'm much more interested in finding out where this is actually useful or where I'm completely wrong than adding another 50 features.
I'd especially like feedback from people who work with:
- precision agriculture
- agronomy
- satellite/remote sensing
- farm management systems
- agricultural data
- smallholder farming
- agricultural research
What would make a system like this genuinely useful on a real farm rather than just being another “AI for agriculture” app?
I'm happy to share the technical architecture or explain how any particular part works.
Satvian: https://play.google.com/store/apps/details?id=com.satvian.app