Tamil Nadu has over 6 million farming families. They grow everything from Ponni rice in the Cauvery delta to turmeric in Erode, bananas in Trichy, and chillies in Virudhunagar. Yet most of them make critical decisions — what to plant, when to sell, whether to take a loan — with nothing more than experience passed down from their fathers.
I built Vivasayi (விவசாயி) to change that. Not as a startup. Not as a product hunt launch. As someone who grew up watching farmers make uninformed decisions that cost them dearly — and who finally had the skills to do something about it.
Vivasayi means "farmer" in Tamil. It is built by a farmer's son, for farmers.
Replace these placeholders with real screenshots from your phone once the APK is built
The Problem I Could Not Ignore
I have seen the numbers. India loses an estimated ₹92,000 crore in agricultural produce every year to post-harvest losses, price crashes, and lack of market information. Tamil Nadu farmers earn, on average, ₹77,000 per year — below minimum wage in most other sectors.
These are not abstract statistics. These are people. And the problems are specific:
These are solvable problems. All it takes is the right information, at the right time — in Tamil.
What Vivasayi Actually Does
Vivasayi is not just another agricultural app with a list of schemes and a weather widget. It is a complete lifecycle companion — walking a farmer from the question "what should I plant?" all the way through to "how do I sell at the best price?"
📱 Login — No Password, Just a Phone Number
Farmers do not remember passwords. I know this because I asked. Vivasayi uses phone OTP login — enter your mobile number, get a 6-digit code, you are in. First-time users set up their farm profile: district, village, land area in acres, soil type (red soil, black cotton soil, alluvial delta soil), and water source (rain-fed, canal, borewell, drip irrigation).
That profile becomes the foundation for everything the app recommends. It is personal from the very first minute.
🤖 AI Crop Advisor — Know Before You Sow
The question every farmer asks at the start of every season: "What should I plant?"
Vivasayi answers it with a local intelligence engine built on Tamil Nadu's actual agricultural data — APMC price history, TNAU crop recommendations, seasonal patterns across all 38 districts. The app scores 19 crop varieties against the farmer's specific soil, water, season, and district — and shows the top crops ranked by profit probability.
A Thanjavur farmer with 3 acres of alluvial soil and canal irrigation, asking in June? The app immediately recommends Samba rice — with a projected profit range and probability percentage, in Tamil, on their screen in 3 seconds.
🌿 Disease Detector — AI Before the Field Officer Arrives
Add your actual disease detection screenshots here
The farmer photographs a diseased leaf. Vivasayi's disease screen uses the Plant.id API — the same AI used by botanists globally — to identify the disease, name it in Tamil, rate its severity (low / medium / high), and give exact treatment steps with chemical names and dosages.
It also works offline. A library of 11 major Tamil Nadu crops covers the most common diseases with reference data stored locally. No internet? No problem. The app still works in the middle of a paddy field in Nagapattinam.
My son showed me how to take a photo of the diseased leaves. The app told me what to spray before I could even reach the agriculture office.
— Farmer, Dindigul📈 Real Market Prices — Crowdsourced by Farmers
Here is something the tech press never talks about: Tamil Nadu mandi data does not exist on India's central data.gov.in API. I discovered this the hard way — after 5,000 API calls across 6 different offsets, Tamil Nadu's total count was zero.
So I built a different solution. Vivasayi's price screen combines two sources:
Seasonal reference prices — calibrated against APMC historical records and MSP data, adjusting automatically by month (harvest season prices differ from lean season prices).
Crowdsourced farmer reports — any farmer can tap "+ விலை பதிவு" and report today's price at their local market. That price immediately appears on every other farmer's screen in the district. Other farmers can "confirm" the price, building a trust score. Green-bordered cards are real farmer reports. Gray cards are reference prices.
This is more useful than any government API. Farmers are the data source.
🌾 Sell Direct — No Middleman, No Commission
A farmer lists their crop: name, quantity, price, district, location description. Takes 60 seconds. That listing is immediately visible to every buyer in Tamil Nadu browsing the app — restaurants, vegetable shops, bulk buyers, export agents.
The buyer taps "Contact via WhatsApp." A pre-filled message opens with the crop details. Farmer and buyer negotiate directly. No broker in between. No 30% commission disappearing.
🌤️ Weather with Farming Advice
The weather screen is powered by Open-Meteo — a free, open API with 7-day forecasts down to 1-hour resolution. But it goes beyond temperature and rain probability. Vivasayi translates the weather data into farming advice:
"கனமழை எதிர்பார்க்கப்படுகிறது. வயல் வடிகால் சரிபார்க்கவும். அறுவடை பணிகளை ஒத்திவைக்கவும்."
(Heavy rain expected. Check field drainage. Postpone harvest activities.)
That sentence is worth more to a farmer than the raw forecast number.
The Growth Stage Lifecycle — The Feature Farmers Love Most
Once a farmer decides what to plant, Vivasayi tracks every stage of the crop's life. Here is what the rice lifecycle looks like inside the app:
| Day | Stage | What the Farmer Does |
|---|---|---|
| Day 0 | Sowing | Transplant seedlings, apply DAP fertilizer at 50 kg/acre |
| Day 7 | Germination check | Fill gaps, first weeding, check water level |
| Day 21 | Tillering | Apply Urea top dressing at 25 kg/acre |
| Day 60 | Panicle initiation | Apply potash, check for blast disease, maintain 5 cm water |
| Day 90 | Flowering | Critical water stage — do not let field dry. Watch for stem borer. |
| Day 125 | Harvest | Drain field 10 days before. Arrange combine harvester. |
Push notifications fire two days before each stage so nothing is missed. The farmer marks each stage done with a tap and can add photos — building a visual diary of their crop's entire growth from seed to sale.
The reminder told me two days before I needed to apply potash. I used to forget this every season.
— Farmer, TiruvarurThe Technology — Built to Work on 2G in a Paddy Field
Every technical decision was made with one constraint in mind: this must work for a farmer in Nagapattinam with a ₹8,000 Android phone on a 2G connection.
Offline first. Market prices, disease guides, weather forecasts, crop recommendations — all cached locally. A farmer can use 90% of the app's features with zero internet connection.
One codebase, three platforms. React Native with Expo means the same code runs on Android, iOS, and web. I shipped for Android first because that is what Tamil Nadu farmers use.
Bilingual by design. Every single screen, label, button, recommendation, and error message exists in both Tamil and English. One tap switches the entire app's language. Tamil is the default because that is the farmer's language.
A hard lesson: Tamil Nadu mandi data is completely absent from India's central data.gov.in API. After testing every offset up to 5,000 records, Tamil Nadu total = 0. The solution was crowdsourcing. Sometimes the government API fails you — and you have to build the data source yourself.
What I Learned Building This
I will be honest about what went wrong as much as what went right.
React Context is broken in this Expo setup. Every attempt at global state via useContext threw "Cannot read property 'useContext' of null." The solution was AsyncStorage-based state — each screen reads its own settings on mount. Not elegant. But it works without crashing.
Plant.id v2 is more reliable than v3. The v3 API returned 400 errors consistently. Switching to v2 with raw base64 (no data URI prefix, modifiers: crops_fast) fixed it immediately.
The best features came from farmers, not specs. I asked farmers in Thanjavur what they actually wanted. "Tell me which crop to plant." That is why the crop advisor exists. I would never have prioritised it from a desk.
WhatsApp is the infrastructure. Every payment, every negotiation, every connection between farmer and buyer — it all happens on WhatsApp. Do not fight it. Build the app to feed into WhatsApp, not replace it.
What Is Coming Next
This is version 1. The vision is much larger.
Version 2 brings OTP login with personal farmer profiles, an AI profit predictor ("If you sow turmeric today, 78% probability of profit at harvest"), push notifications for price alerts, and a full season tracker from sow date to harvest.
Version 3 integrates UPI payments so buyers can pay farmers directly inside the app, logistics connectors for last-mile delivery, FPO (Farmer Producer Organisation) support for collective selling, and government scheme eligibility checking.
The long-term goal is simple: a Tamil Nadu farmer should be able to open Vivasayi at the start of a season and have everything they need to make the right decisions — what to grow, when to spray, what disease is killing their crop, and who will buy it at a fair price — all in one app, in their language, offline.
Try Vivasayi — It Is Free
Download the Android APK and install it on any Android 8.0+ phone. No subscription. No login required for most features. No ads. Ever.
Android 8.0+ · 25 MB · Tamil + English · Works offline
Install instructions: Settings → Security → Allow unknown sources → Open downloaded APK → Install
Built with React Native · Expo · Supabase · Plant.id AI · Open-Meteo
Tags: Tamil Nadu · Agriculture · React Native · Farmers · App Development