Inspiration I saw my brother, a hardworking farmer, struggling every season.
He didn’t know the right time to sell his crops, and sometimes he sold them at very low prices. He found it hard to rent machines or arrange transport, and since he isn't fluent in English or tech, using most farming apps was frustrating for him
What it does It’s a voice-enabled AI assistant that helps farmers with everything they need — in their own language.
🌾 Key Features: 🎙️ Voice Assistant in Local Language Farmers can talk to the app to ask about prices, markets, tools, and transport — no need to type.
📈 AI-Based Price Prediction The app predicts the best time and price to sell crops based on market trends and location.
🛒 Produce Marketplace Farmers can list their crops for sale and connect with buyers directly — without middlemen.
🚜 Equipment Rentals Need a tractor or sprayer? The app helps farmers find and book nearby machines.
🚚 Logistics Booking Farmers can book transport to deliver their produce to mandis or buyers.
How we built it Frontend (User Interface)
Built a mobile app using React Native for cross-platform support (Android/iOS).
Designed a simple UI with voice input so farmers can speak instead of typing.
Added multi-language support (starting with regional Indian languages).
Voice Assistant Integration Used Google Speech-to-Text API or OpenAI Whisper to convert spoken language to text.
Used Text-to-Speech (TTS) so the app could reply in the farmer’s language.
Created custom voice commands for different actions like “sell crops,” “check price,” etc.
AI Price Prediction Trained a basic machine learning model (e.g., using Python + scikit-learn or TensorFlow) on:
Crop type
Historical prices
Season/time
Location
The model gives farmers a predicted best price range. Backend & Database Set up a Node.js or Python Flask backend with a Firebase or MongoDB database.
Stores user data, listings, rental records, and transport bookings.
APIs & Services Integrated:
Agri market APIs for real-time crop prices.
Map APIs for locating rental equipment and logistics services.
Authentication APIs for secure farmer login.
Hosting & DevOps Deployed the backend on Render / Vercel / Heroku or AWS.
Used GitHub for version control.
Planned for Docker-based deployment (optional DevOps extension).
Challenges we ran into Language & Voice Understanding
Making the app understand regional languages and dialects accurately was tough. Many farmers speak mixed or non-standard forms of language, which made voice recognition challenging.
📶 Rural Connectivity Issues We had to ensure the app works in areas with low or unstable internet. Optimizing it for slow networks took time.
🤖 Accurate AI Predictions Training an AI model to predict crop prices with accuracy required a lot of historical data, which was hard to find and often incomplete.
🧑🌾 User Interface Simplicity Designing an app that even non-tech-savvy farmers could use was difficult. We had to test and rework the UI to make it simple, voice-driven, and intuitive.
📦 Building a Reliable Marketplace Creating a system where farmers, buyers, and logistics providers can safely interact took effort — especially around trust, safety, and verification.
🔐 Data Privacy & Security Ensuring the farmers' data (location, crop info, personal details) was secure while using cloud services was also a key concern.
Accomplishments that we're proud of Built a Fully Voice-Enabled Assistant
We successfully created a farming assistant that works with voice commands in local language, making it accessible to rural users who are not tech-savvy.
📈 Integrated AI Price Prediction Our app can now predict crop prices using machine learning, helping farmers make smarter selling decisions.
📱 Simple, Farmer-Friendly UI We designed an interface so simple that even first-time smartphone users can navigate it with ease.
🚜 Enabled Real-time Equipment & Logistics Booking Farmers can now rent tools or book transport directly through the app — something that usually required phone calls or middlemen.
🌾 Created a Mini Agri Marketplace We built a platform where farmers can list and sell their produce directly to buyers, cutting out intermediaries.
🧪 Tested with Real Users We got early feedback from actual farmers and improved the app based on their needs — a big step toward real-world impact.
What we learned Listening to Real Users is Everything
Talking to farmers helped us realize that simplicity matters more than features. We learned to build for people, not just technology.
🌐 Building for Rural India Requires Different Thinking We had to rethink how apps work in low-connectivity areas, using lightweight design and offline-first ideas.
🗣️ Voice is More Than a Feature — It’s a Lifeline We learned how powerful voice interfaces can be for people who don’t read or write in English — it creates true digital inclusion.
🤖 Training AI Needs Clean, Local Data AI price prediction sounds great, but we learned that getting quality agricultural data — especially at the village level — is a huge challenge.
💡 User-Centered Design Makes All the Difference Simple buttons, local language, large icons — these small things helped us design a more usable and meaningful app.
🤝 Collaboration is Key Working across AI, design, voice tech, and backend taught us how important teamwork and communication are during a fast-paced build.
What's next for Untitled Expand Regional Language Support
We plan to add more Indian languages and dialects to make the app accessible to farmers across the country.
📶 Offline Mode + SMS Integration We're working on enabling offline functionality and SMS-based alerts, so farmers can still use the app without internet.
🤖 Smarter AI with Real-Time Market Feeds We aim to integrate live mandi prices and weather data to make our AI price prediction even more accurate.
👥 Partner with FPOs and NGOs Collaborating with Farmer Producer Organizations and agri-NGOs will help us reach real users faster and scale responsibly.
🔒 Secure Farmer Identity & Verification We’ll build systems for user verification, fraud prevention, and safe transactions in the marketplace.
📲 Launch on Play Store Final goal: Launch a pilot version on the Google Play Store and test it in selected rural areas.
Log in or sign up for Devpost to join the conversation.