Kisan e-Mitra: India’s voice-enabled AI support for farmers

The problem

India's Pradhan Mantri Kisan Samman Nidhi (PM-KISAN) is a national government scheme that provides regular cash instalments directly to farmers who own land. It was launched in February 2019 to provide additional financial support to landholding farmers, and it operates at a very large scale across the country.

Farmers with a question about the scheme, whether they qualify, how to register, when a payment is due, or why an instalment has been delayed, had few reliable avenues to get an answer. They typically had to travel to a district agricultural office or to a Common Service Centre, one of the access points set up under India's Digital India programme to deliver government and public services in rural and remote areas. For many people in rural areas, a single such trip could absorb a day's earnings. Official guidance was often available only in Hindi or English, which left farmers who speak one of India's many other languages at a disadvantage; the country has 22 officially scheduled languages. Limited familiarity with technology and literacy difficulties further narrowed the pool of people who could use the existing channels, compounding the extended waiting periods for queries and complaints to be resolved.

The solution

Closing the gap between farmers and the scheme meant giving people a way to get answers directly, without travelling to an office or relying on an intermediary. To do that, the Ministry of Agriculture and Farmers' Welfare introduced Kisan e-Mitra, an AI chatbot. It was launched in New Delhi on 21 September 2023 and described at launch as India's first AI chatbot connected to one of the national government's major schemes.

Farmers reach it through the PM-KISAN website and app. They can ask in their own language, by speaking or typing, about matters such as eligibility, registration, payment status, or why an instalment has not arrived. The chatbot uses natural language processing, the branch of AI that lets software interpret everyday human language, to interpret each query and return information specific to that farmer.

Coverage across languages is provided through Bhashini. Bhashini is the Indian government's national platform for language translation, launched by the Ministry of Electronics and Information Technology, designed to enable citizens to access digital services in their own languages. The chatbot first launched in five languages: English, Hindi, Bengali, Odia and Tamil, and is now accessible in 11 regional languages, with the stated aim of eventually covering all of India's 22 officially scheduled languages.

The impact

The chatbot's effect shows in two registers: the volume it now handles, and the access it opens for farmers previously held back by distance, language or literacy.

1. Tens of thousands of queries a day, millions in total

Government reporting in April 2025 recorded over 20,000 farmer queries handled daily and more than 92 lakh, or 9.2 million, queries answered in total.

2. Nationwide and around the clock

The service runs 24/7 across all 28 states and 8 Union Territories, the regions the national government administers directly. More than 400,000 farmers are enrolled, and over 290,000 have been served.

3. Voice enablement increases accessibility

In rural areas, literacy varies, and farmers may feel more at ease speaking than typing, so accepting spoken questions makes the service usable by people approaching a digital government tool for the first time, without the barrier of typing a question. This widens who can use the scheme's information directly: a text-only service would have depended on a farmer being able to read and write a query, leaving those with limited literacy reliant on someone else to do it for them. Spoken access lets them reach the same answers themselves, in the language they speak.

4. Positioned as a transparency measure

The Ministry of Agriculture has presented the tool as part of efforts to make scheme delivery more transparent. The reasoning is that a farmer who can check their own payment status or application progress relies less on local intermediaries, which the Ministry links to less room for misinformation and local-level delays. Farmers can get answers in their own language without going through an intermediary or waiting in a queue.

Beyond PM-KISAN, the Ministry has indicated plans to integrate the chatbot with other major government schemes that empower farmers, including critical information on fertiliser subsidies, minimum support prices, and weather advisories.

Lessons learned

• An existing national language platform handled the multilingual side. Because Bhashini, the government's shared language service, was already running, the chatbot did not have to solve translation and speech across Indian languages from scratch; it drew on a platform built for that purpose. For a government weighing whether to deploy a similar tool, the existence of strong language infrastructure minimises the need to build translation capabilities into each individual initiative.

• The PM-KISAN scheme's existing records enabled farmer-specific responses. Kisan e-Mitra could tell an individual farmer their payment status or application progress because PM-KISAN already held structured, per-beneficiary records and an online portal for the chatbot to draw on. Where that kind of organised, queryable data is already in place, an AI tool is better able to provide clear and specific assistance.

This case study was written with assistance from artificial intelligence.