Inspiration
The inspiration for AgriAI comes from the common issue faced by farmers in Pakistan—unfair pricing and exploitation by middlemen. Farmers often sell their produce at low prices to industrialists who then sell it to buyers at much higher prices, leading to financial loss for the farmer. Our goal was to create a direct connection between farmers and buyers, ensuring farmers get fair value for their products.
What it does
AgriAI is an AI-powered marketplace application that connects farmers directly with buyers. The app allows farmers to list their products via voice commands in Urdu, eliminating language barriers and simplifying the process for those with limited literacy. The platform also includes a bidding system, where farmers can set their price and buyers can bid, with the maximum price bound by government-regulated prices.
How we built it
We developed AgriAI using a combination of technologies:
- Voice Recognition AI to convert speech to text in Urdu for farmers who may not be familiar with reading or writing.
- Bidding System where the initial price is set by the farmer, and buyers can place bids up to the government-set price limit.
- Marketplace Platform using mobile-first design to ensure accessibility for farmers in rural areas.
- Backend powered by scalable cloud services to ensure smooth operation and real-time updates.
Challenges we ran into
- Voice Recognition in Urdu: Building a robust speech-to-text AI that could accurately understand farmers’ regional accents and dialects was a significant challenge.
- Connectivity Issues: Many farmers in rural areas have limited access to stable internet, which made developing a solution that works offline or with low bandwidth a priority.
- Government Price Data Integration: Integrating real-time updates for government-set prices was challenging due to the lack of a centralized database for agricultural pricing.
Accomplishments that we're proud of
- AI Voice Agent: We successfully developed a speech-based system in Urdu, which helps farmers list their products and communicate with buyers.
- Bidding System: The bidding system is fully functional, ensuring farmers can set prices with transparency and fairness.
- User Accessibility: We designed a user-friendly interface that allows farmers with minimal tech knowledge to easily interact with the app.
What we learned
- User-Centric Design: We learned the importance of designing for the user, especially when working with marginalized groups who may not be familiar with technology.
- Overcoming Technical Limitations: Developing a speech-to-text system for Urdu required overcoming technical barriers, and we gained valuable insights into AI and machine learning applications for regional languages.
- Understanding Rural Connectivity: The challenges of internet connectivity in rural areas taught us to focus on making our app lightweight and capable of functioning with minimal data.
What’s next for AgriAI
- Expansion of the Marketplace: We aim to expand the platform to more farmers across Pakistan, ensuring that it’s accessible to even the most remote regions.
- Integration with Financial Services: We plan to integrate payment gateways and financial services to allow farmers to receive payments directly through the app.
- Enhanced AI Capabilities: We are looking to improve our AI’s accuracy and add features like product recommendation, dynamic pricing, and market analytics for both farmers and buyers.
Log in or sign up for Devpost to join the conversation.