When black patches began creeping up the stems of his mango saplings, Shiv Kumar Yadav knew something was wrong.
The four-acre nursery owner from Uttar Pradesh’s Varanasi district watched as white fungal growth spread across the young plants and healthy leaves began dropping one after another. “I did not know what disease it was, but I knew I had to act quickly,” the farmer recalls.
Like countless farmers across India, his first instinct was to seek advice from local pesticide dealers and neighbouring farmers. But the advice was either delayed or too general to solve the problem. By the time farmers receive the right guidance, the damage is already done.
Shiv refused to leave it to chance.
Instead of trying another pesticide based on guesswork, he reached for his smartphone. He photographed the infected saplings, uploaded the image to an application, and described what he was seeing. Within seconds, he received a diagnosis along with clear instructions on how to tackle the problem. He removed the infected plant parts, improved drainage, avoided overwatering, and applied the recommended fungicide in the prescribed quantity.
“Within a few days, the spread of the disease was under control. It helped me save a large number of saplings and reduced my losses,” he says.
For India’s more than 150 million farmers, decisions like these cannot wait. A sudden pest attack, an unfamiliar disease or an unexpected spell of rain can alter the course of an entire season.
Recognising this gap nearly two decades ago, ‘Digital Green India’ has spent years trying to make sure that expert agricultural knowledge reaches farmers when they need it most. Its latest effort, a free artificial intelligence-powered application called ‘FarmerChat’, has already reached nearly one million users across the country, helping farmers make faster and more informed decisions in the field.
When information exists but never reaches the field
India has no shortage of agricultural knowledge.
Research institutions continuously develop improved farming practices. Agricultural universities study pests, diseases, and climate-resilient cultivation techniques. Government departments regularly issue advisories. Despite this vast pool of expertise, many smallholder farmers still struggle to access practical guidance when problems arise.
Agricultural universities study pests, diseases, and climate-resilient cultivation techniques.
The challenge, says Digital Green India CEO Nidhi Bhasin, has always been about delivering information at the right moment.
“The gap in Indian agriculture was never just that information did not exist,” she tells The Better India. “The challenge was guaranteeing it reached farmers in the right language, at the right time and in a way they could actually use.”
Launched in March 2008, the organisation began by strengthening the country’s existing agricultural extension system rather than replacing it. The non-profit worked with state governments, State Rural Livelihood Missions, agricultural departments, and grassroots organisations to create farmer-to-farmer learning videos teaching practical farming techniques in local languages.
Over the years, those efforts expanded across Bihar, Jharkhand, Odisha, Andhra Pradesh, Telangana, Maharashtra, Karnataka, Madhya Pradesh and Uttar Pradesh. Today, the organisation has produced more than 10,000 agricultural videos and reached around 6.5 million farmers and extension workers through its various initiatives, with women forming the majority of participants.
As internet access improved across rural India and smartphones became common, the organisation realised it could take the same trusted knowledge directly into farmers’ pockets.
That led to the launch of FarmerChat in October 2024.
A new chapter led by technology
Leading this progress is Nidhi Bhasin, who joined Digital Green India as chief executive officer in September 2024 after spending more than 25 years in the social development sector.
“I had spent years understanding social impact and years understanding how technology can create change,” she says. “This space allowed me to combine both and solve problems where they matter most. Agriculture supports millions of livelihoods, so even small improvements can create a very large impact.”
Why FarmerChat was created
The idea behind FarmerChat grew out of a simple observation.
Farmers need immediate advice. Whether deciding when to irrigate, identifying a pest, selecting seeds, or responding to changing weather, every farming decision depends on timing. Waiting several days for an agricultural officer to visit can mean losing an entire crop.
Climate change has only made those decisions more urgent.
Erratic rainfall, rising temperatures, and unfamiliar pest outbreaks have made farming increasingly unpredictable, leaving smallholder farmers searching for reliable information that can help them respond immediately.
“That is exactly the gap this platform was designed to fill. Farming is incredibly time-sensitive. A farmer may need to decide today whether to irrigate, spray, or protect a crop from changing weather. Waiting several days for advice can mean devastating losses, which is why we wanted to put timely, reliable guidance directly into a farmer’s hands, in a language and format they are comfortable using,” says Bhasin.
Today, the organisation has produced more than 10,000 agricultural videos and reached around 6.5 million farmers.
Unlike many digital platforms that provide generic information, this app was designed to offer personalised guidance based on a farmer’s specific situation.
The application is available free of cost on Android through the Google Play Store. Farmers can either type a question, record a voice message, or upload a photograph of their crop. The platform currently supports English, Hindi, Marathi, Kannada, Telugu, and Odia.
“It is an AI-powered agricultural assistant in the farmer’s pocket,” she explains. “It is designed to take a farmer from confusion to action. Whether someone asks a question through voice, text, or an image, they receive advice in the same language and format they are comfortable with.”
Making the platform free for farmers has been a conscious decision.
Because it is a nonprofit organization, it does not generate revenue from the app. Instead, the costs of building, maintaining, and expanding the application are supported by philanthropic organizations, corporate social responsibility partnerships, and ecosystem collaborators.
“Our objective has never been to charge farmers for knowledge. Support from philanthropies and CSR partners allows us to keep the app completely free while continuing to improve the platform and make it accessible to more farmers,” says the CEO.
From a photograph to a solution in seconds
A farmer can type a question, record it as a voice message, or upload a photograph of the crop. But behind that simple interface is a process that brings together several layers of information before arriving at an answer.
Once a query is submitted, the platform first understands the farmer’s location, crop, stage of cultivation, and local weather conditions. It then searches a verified agricultural knowledge base built using government advisories, scientific research, and Digital Green’s own repository of farmer-led learning collected over nearly two decades. Only after combining this information does it generate a response in the farmer’s preferred language, either as text or as an audio message.
“Our aim is not just to answer a query. We bring together local conditions, crop information and verified agricultural knowledge so farmers receive advice that is timely, reliable and useful in the field,” she explains.
The application supports advisory on crops, livestock, dairy, fisheries and climate-smart farming practices, but two kinds of questions consistently dominate.
When every leaf tells a story
Crop diseases rarely announce themselves.
The first signs are usually easy to miss. A few spots on the leaves, stems that begin to darken, or fruits that fail to develop properly. Left unchecked, those small changes can rapidly spread across an entire field.
Earlier, farmers mostly relied on guesswork. Some carried affected leaves to local pesticide shops. Others sought advice from neighbours or waited for agricultural officers to visit.
The platform shortens that entire process.
A farmer directly photographs the affected crop and uploads the image. If more information is needed, the application asks follow-up questions about the crop or location before analysing the symptoms against weather conditions and verified agricultural data.
Within moments, it identifies the likely disease or pest attack and recommends suitable treatment. Farmers receive both chemical and organic options, along with the exact quantity to use, the method of application, and additional precautions.
If doubts remain, they can continue the conversation.
“Farmers ask us how much fungicide they should use or whether they should spray again after a few days,” says the CEO. “The app allows them to keep asking questions until they are confident enough to take action.”
That guidance changed Shiv Kumar’s nursery. “Earlier, delayed diagnosis meant diseases spread quickly and I lost valuable saplings. Now I can identify problems early, take action immediately, and avoid unnecessary spending,” he says.
Over one season, he estimates he saved around Rs 10,000 by reducing plant losses and avoiding repeated pesticide applications.
Unlike many digital platforms that provide generic information, this app was designed to offer personalised guidance based on a farmer’s specific situation.
Another farmer, Vishwesh Kumar Singh from Babiuan village in Varanasi, experienced something similar after yellow and brown spots appeared across nearly an acre of his wheat field.
He uploaded the symptoms through the app. The platform recommended carbendazim and hexaconazole in the correct dosage while also advising him on irrigation and field monitoring.
“I avoided repeated pesticide applications and used only the quantity that was actually needed,” he says. “That reduced my cultivation costs and prevented the disease from spreading further.” He estimates the timely intervention helped him save between Rs 8,000 and Rs 10,000 during the season.
Sometimes the best advice is to wait
Not every farming decision involves treating a disease.
Many revolve around timing. Should fertiliser be applied today or after the next rainfall? Should irrigation happen now or can it wait? Is this the right week to sow potatoes?
For generations, farmers have relied on experience, local observation and conversations with neighbours to answer these questions. But unpredictable weather has made those decisions far more difficult.
Instead of offering one standard recommendation, the platform studies the farmer’s location, crop stage, weather forecast and advisory datasets before suggesting whether fertiliser should be applied immediately, split into smaller doses or postponed altogether.
Bhasin recalls the example of Anu Kumari, a farmer from Bihar who was preparing to sow potatoes when dark clouds began gathering overhead.
Instead of relying on instinct, she turned to the app. It predicted rainfall and advised her to postpone sowing. Heavy rain arrived the very next day. “Had she planted that day, much of the crop would have been washed away,” she says. “By waiting, she avoided a loss of nearly Rs 5,000.”
A similar experience changed the way Urmila Devi, a smallholder farmer from Mirzapur, manages her wheat and paddy fields. She first came across the platform during a Digital Green meeting in July 2025 while searching for reliable advice on fertiliser application and irrigation.
During the wheat season, the application recommended balanced fertiliser based on the crop’s growth stage, suggested the right time for irrigation, and advised her on weed and pest management.
“It helped me use only the inputs I actually needed, and compared to the previous season, I earned around Rs 7,000 to Rs 9,000 more from my wheat crop,” she says.
Teaching AI to stay grounded in reality
Artificial intelligence can generate impressive answers, but agriculture leaves little room for mistakes. One incorrect recommendation can damage an entire crop.
That is why the organisation built FarmerChat differently from a general-purpose chatbot.
Making the platform free for farmers has been a conscious decision.
The platform combines large language models with retrieval-augmented generation, commonly known as RAG. Instead of searching the open internet or inventing responses, it generates answers only from a carefully verified agricultural knowledge base. It also uses smaller, fine-tuned language models, voice technologies for Indian languages, and image recognition to identify crop diseases from photographs.
Technology, however, is only one part of the process. Human expertise remains the core.
Across different states, agronomists regularly review farmer queries, validate responses, and continuously improve the system using real farming situations. In Bihar alone, four to six agronomists helped train one of the platform’s language models before it was released.
“Technology should support experts, not replace them,” says Bhasin. “We have built a human-in-the-loop system where agricultural specialists continuously validate and improve the advice so farmers can trust what they receive.”
Nearly three in four farmers rated the app’s advice as highly relevant, while four in five said it provided all the agricultural information they needed.
Building confidence alongside crops
FarmerChat’s rapid adoption has surprised even its creators. In less than two years, the application has reached nearly one million users across India.
The CEO believes the growth has been driven by three things. Trust, accessibility and relevance. “We were not building an app in isolation. We had already spent years listening to farmers and understanding the problems they actually wanted solved,” she says.
Today, almost half the application’s users are women, a number that continues to grow.
Because the platform accepts voice queries and replies in local languages, literacy is no longer a barrier to accessing expert advice.
According to an independent study by 60 Decibels, 96 percent of women users reported greater confidence in making farming decisions after using the app, compared with 89 percent of men. That confidence is reflected in farmers such as Jyoti Devi from Mirzapur.
Her bitter gourd crop once produced small fruits that rotted before harvest.
“Even after growing bitter gourd, I had to buy it from the market for my own family,” she says.
After following FarmerChat’s recommendations on drainage, nutrient management and crop protection, her harvest improved dramatically.
Today she harvests between 20 and 30 kilograms of bitter gourd every week, earns nearly Rs 6,000 a month from surplus sales and no longer buys the vegetable for her household.
“We now eat vegetables from our own farm and even share them with our neighbours,” she says.
For Digital Green, these stories matter more than download numbers.
Independent studies found that 60 percent of active users acted on the platform’s advice, 83 percent said agricultural information had become easier to access, and 91 percent reported greater confidence in making farming decisions.
Looking ahead, the organisation hopes the app will become much more than an advisory platform. Future versions aim to support personalised crop calendars, government scheme discovery, market prices, credit support and market linkages, creating a single ecosystem where farmers can access solutions throughout the agricultural cycle.
A farmer can type a question, record it as a voice message, or upload a photograph of the crop.
As climate change continues to make farming more uncertain, the CEO believes responsible AI will become increasingly important. “AI can become a practical resilience layer for Indian agriculture,” she says. “If it is built responsibly, it can make real-time, localised knowledge available to every farmer, regardless of gender, literacy or location.”
For Shiv Kumar, that future has already arrived. The next time something unusual appears on his mango saplings, he knows exactly where he will turn first. Not because technology has replaced his years of farming experience, but because it has made sure that reliable advice is never far from his pocket.
All pictures courtesy Digital Green India.
