Khetiyaar Case Study: Designing a Free AI Farming App for Indian Farmers in Three Languages

Khetiyaar is a free-to-start Android app that puts an AI farm expert, a season-long crop planner, weather, mandi rates and a local marketplace in one place, in Gujarati, Hindi and English. This case study covers the problem it set out to solve, the design decisions that shaped it and what other founders building for non-technical, multilingual users can borrow.

The challenge: one season, many tools

A farmer's season is a chain of decisions: what to sow, when to water, what a leaf spot means, where to sell and whether the season made money. Most apps solve one link. Farmers end up juggling a weather app, a messaging group for advice, a mandi site for prices and a notebook for costs. The brief was to join those links into one daily habit without asking users to learn a new vocabulary.

Design decisions that mattered

1. Language first, not translation last

Gujarati, Hindi and English are treated as full citizens. The site gives each language its own URL and connects them with hreflang, so a farmer searching in Gujarati lands on Gujarati pages, and the app and library follow the same rule. Translating a finished English product almost always leaves awkward strings in the corners; designing in three languages from the start avoids that.

2. Big buttons, plain words, easy at 50+

The interface leans on large touch targets, short labels and one clear action per screen. The aim is that someone using a phone mainly for calls and WhatsApp can still find today's task. Voice input for the AI chatbot lowers the typing barrier further, and learning guides can be read aloud.

3. Coins instead of a subscription

AI costs money every time it runs, and farm income is seasonal. Rather than a monthly fee, Khetiyaar uses prepaid coins: new farmers get a welcome bonus, daily chats are free, the first diagnosis is free, prices are shown before every spend, and coins are refunded automatically when the AI fails on the app's side. Coin packs are bought through Google Play billing and are clearly described as in-app credits, not money.

4. Honest advice over big promises

Agriculture advice can cost a harvest, so the product voice avoids guarantees and shows the limits of what the AI can say. That trust is also a growth lever: the Khetiyaar team publishes a candid comparison with other farming apps, including where Khetiyaar falls short, which is the kind of page answer engines like to cite.

What is inside the app

• AI chatbot: a 24x7 farm expert by voice or text.
• AI crop planning: a day-by-day season plan.
• AI garden planning: 13 methods for balcony, terrace or kitchen growing.
• Weather: humidity, wind, soil temperature and moisture, plus daylight.
• Land measurement: walk the field and get the area in acre, bigha or vigha.
• Farm calculator: log every cost and see where the money went.
• Mandi bhav: today's market rates and which market pays more.
• Agri Bazar: buy and sell animals, crops and equipment, and see what buyers want.
• Farming Pulse: live, area-wise trends of what nearby farmers plan to grow.
• Learning library: 500+ guides that read aloud.

Built to be found: SEO, GEO and AEO

The product launched with the search groundwork in place. Each language has its own canonical URL and hreflang alternates, the home page carries Organization, MobileApplication and FAQPage structured data, and the FAQ answers lead with a direct one-paragraph answer, the format AI answer engines quote. Dedicated landing pages cover the searches farmers actually make: crop planning app, farming chatbot, mandi bhav, gardening app, organic farming app. Free tools — a farming calculator and a land converter — give people a reason to link and return.

What we would do next

• Publish real outcome data — time saved, plans completed, prices checked — once users approve it. Until then we only claim what the product does.
• Extend beyond Android when the audience demands it.
• Add more regional mandi coverage and content in additional languages.

Lessons for your own app

1. 1.Design for the least technical user in your audience, then everyone else benefits.
2. 2.Make costs visible before they happen. Trust is the product.
3. 3.Treat each language as its own product surface for design and search.
4. 4.Put your honest limits in the open; it differentiates you in AI answers.