this startupxa0built a treasure trove of crop data by putting a.i. in the hands of indian farmers

Tushar Kamble knew that his peppers were smaller than ordinary ones, that the leaves began to curl, and that he knew there was a problem with the peppers.
Last year, the 30-year-old former scholar boldly threw himself into agriculture, and one morning, on the border between his 7-acre land and their land, he asked the neighboring farmers for advice.
Kamble, 38, has many different views on bees and diseases.
He then tried an app called ledplantix and used it to take photos of his Chilean plants. The app cross-
Using image recognition, make reference edits to 50 different species in the database
Within two minutes he had a different answer: his children did not get enough water and they would benefit from the trace nutrient spray.
In a few weeks, Kamble\'s peppers have grown into a decent size.
Today he is using the app to help him change his chemical use so that he can have the opportunity to export the crop to Europe.
\"I don\'t rely on other people to help me,\" he said with a smile . \".
Raising a farmer\'s crop production seems trivial, but small farmers like him make up the long tail of global food production.
About 70% of the world\'s food comes from small farmers like Kamble, with about 0. 5 billion worldwide, said Simon Strey, founder and CEO of plan.
Her company is using pictures from Kamble and other companies to have a more comprehensive understanding of crop health in India and other regions.
Strey started with her husband and her machine.
In 2015, study expert Robert Strey had 620,000 monthly active users, 80% of whom were in India.
Sandip Shinde is another farmer who uses Plantix on 25 acres of land where he grows tomatoes, broccoli and sugar cane, although he has 13 years of experience, he believes, apple\'s automated advice system and chat networks with other farmers have also helped him boost production.
But what is good for small farmers can also be useful for the world\'s largest agricultural and chemical companies.
Peat found the connection between the two worlds in the original business model, white-
Label the factory
BASF Chemical\'s identification software, which has integrated it into its own crop management software, Maglis.
Professional farmers using Maglistend have much larger fields than Indian small farmers, and they also visit Peat\'s software more often --
Up to 12,000 times a month.
Startups charge for each \"call\" of their API, or every time a professional farmer uses Peat\'s features on the Maglis app.
But Peat can also use images of these big farms to help train images of it
Recognition algorithm.
This may be more valuable than the income stream (
According to Strey, it is still far below $1 million a year).
All this helps to build predictive models of disease transmission on crops from wheat to potatoes to pomegranates, as well as possible effects on food production.
She has a data scientist on her team, and more employees join in, the former geographer and soil scientist Strey plans to sell these predictive insights to governments, insurance companies and agricultural suppliers, and the latter two are \"the most promising stakeholders to sell data [to].
\"When we talk to seeds and plants --
They said, protecting the supplier \'We don\'t have data on what they need, \'\"said Strey of the small farm.
She added that the same is true for farmers in Africa and Asia, most of whom are grown on less than two hectares of land, or on the size of a football field.
\"There is no database for these farmers.
In the fall of 2014, when steely and her husband were doing soil research in Brazil, they had a preliminary idea of clay coal.
A farmer in a village mentioned that whenever he tries to search Google for a plant disease known as \"sudden death\", it will only have images of a car accident, nothing to do with the crop.
Since the local language was a clear barrier, the couple thought it would be easier for farmers to take photos of crops and get an automated treatment answer. Today Berlin-
Headquartered in Peat with 25 employees and raised nearly $5 million in venture capital
In December 2017, their latest round A financing was led by London index investment.
Strey\'s application, if it can continue to spread across the world of small farmers in the developing world, shows the growing popularity of artificial intelligence or machines --
Learning tools can turn a lot of data into something valuable.
As a form of artificial intelligence, image recognition once surpassed the scientific field.
Fi, but thanks to off today-the-shelf machine-
Free learning libraries from Google, Microsoft and even Amazon and Facebook, the behavior of \"using artificial intelligence\" is becoming commercialized, benefiting unique, difficultto-
Data like Peat.
Peat uses Google\'s TensorFlow software library for its images
The recognition tool, noted Strey.
\"Everyone can do this,\" she said of her tools . \".
\"What we think is ultimately the most powerful is to create insights from the data we get from our users.