Food Wastage
Perfectly edible food is wasted at restaurants daily.
Solution:
Predict sales of food items so as to prepare the right amount for a day. Reduces wastage. Increases profit
Prediction algorithm based on Previous sales data and variables like season, events or festivals happening and their scale
Additional Features - Generates a list of ingredients required to make predicted food items. Suggests these from ONDC, hence avoiding middleman and helping in profitable deals for both farmers and restaurants.
This was the idea mentioned in the abstract submitted. (The name PlatePal was changed because there is a vehicle number plate registration company of the same name.) This is implemented. A few additional features are mentioned below.
Lack of Modernization in agriculture
Lack of knowledge on advancements in agritech - Most farmers only know the regional languages. A specially trained voice chat bot in regional language for farmers to learn about new technological solutions to their issues.
Lack of resources - Most farmers lack resources to buy new and expensive tech. The solution is to pool resources. A farmer can propose a pool request to buy a particular tech solution and others can contribute to it. The money will be held in the company's account and only after all farmers agree, the total amount is transferred to the pool initializer. This is to prevent theft attempts.
Rental Services - Farmers can rent their tools to other farmers and earn a small profit from their otherwise idle resources.
Huge Middlemen margin
Middlemen exploit farmers and take huge margins for products. Integrating selling through ONDC helps in avoiding middlemen and providing an even marketplace for small retailers like small landholder farmers. Farmsta provides this kind of marketplace
ONDC platform integration. No tutorials and the documentation seems to be confusing and long. Also there is no sdk or trial api making it difficult to integrate it into new products. The dev sandbox also requires credentials which are hard to obtain.
Challenges with packages in flutter. Some packages create huge audio files creating issues in voice chat. Trying a different package solved the issue.
Translation to regional language - Tried whisper, but it wasn’t compatible with malayalam. So google translate api was used.
Tracks Applied (3)
Discussion