Agriculture today is driven by multiple sociological, economic and environmental forces.

Farmers today live in an environment that is in a constant state of flux and are being challenged to react and adapt at a pace seldom seen before.

Amongst a host of scientific developments and adoptions we at Veeville.farm see AI as the most empowering technology yet in the way we grow and manage our crops. Our technologies allow farmers to gain a better understanding of the “ground realities now” more accurately and more quickly than ever before.

We are developing an Advisory Service to answer questions and clear doubts, so the farmer only grows crops and not questions.

And unlike a typical adviser the farmer can have a conversation with the Bot.

The engine powering the advisory service also initiates recommendations that are specific to that particular farmer based on hyper-local and global considerations.

The advisory engine is based on a state of the art Machine Comprehension Model that can digest content – like recorded best practices, chat logs etc and answer specific questions.

So the farmer can ask anything from “what the irrigation time table is for typeX potatoes” to “What is the min distance between seeds when planting potatoes?” and the Bot will answer.

“We have built an AI pipeline that can classify images, audio and videos for the content they have and segment them for response.”

We have enabled the platform to be as natural to use as possible and one of the most important ways to interact is the ability to just talk to the bot.

We have designed speech recognition and designed it to suit this particular use case with special terms recognition etc.

So when the farmer talks to the Bot asking “How much many kgs of fertilizer for 3 acres of BT1023” the agent knows exactly what she is talking about!

“AI lets the farmer diagnose diseases, identify weeds and bugs, classify the state of the crop etc. even when he is off-line”

There may be situations where the farmer wants to know something specific about what he is seeing, so we built a multi-modal model to let him do just that.

The farmer takes a picture and ask a specific question about it and the Bot will try and answer that. This also works for processes, so when the farmer point his camera at a new tool or pesticide and asks “How do I do this?” the Bot knows exactly what he wants.

Or he points his camera at a fruit and asks “Is that ripe for harvest?” the bot can analyze the image and give him an answer.

“Our FarmKit combined with the Multi-modal AI models enable the farmer to get unprecedented insights of the process and enables fine grain control of the parameters that affect the crops.”

The Workflow feature enables the farmer to shift the heavy lifting of tracking the steps in a specific process to the Bot. The Bot can keep track, raise flags and send notifications etc. And with connected devices the APIs also enable talking to other connected Farm devices.

So, for instance, setting the sprinklers to turn on when the humidity falls below a certain threshold is now a breeze.