Wagabit
Wagabit connects buyers with relevant sellers or services locally.
What it does
Wagabit is a platform that connects buyers with relevant sellers or service providers in their local area. It allows users to submit requests for products or services they are looking for, specifying details and setting a preferred location radius. Sellers and service providers within that radius are notified of the requests and can respond through the Wagabit platform, initiating contact with the buyer. This ensures that buyers receive offers from vendors that are geographically convenient and tailored to their needs, while sellers gain access to a targeted audience actively seeking their products or services.
The Gemini API is utilized within Wagabit to efficiently match buyer requests with potential sellers. Seller profiles, including detailed descriptions of their products and services, are saved in Pinecone, a vector database. When a buyer submits a request, the system first filters the seller data using Pinecone’s vector search, comparing the buyer’s request against the stored seller vectors to find the most relevant matches. This initial filtering narrows down the potential matches based on similarity to the buyer's request. The filtered results are then submitted to the Gemini API for a final, refined match that considers additional criteria such as location and specific preferences. This two-step filtering process ensures highly accurate and relevant connections between buyers and sellers, facilitating prompt and successful transactions.
Built with
- In addition to the Gemini API
- I used Google Maps API for geolocation services
- allowing users to set their preferred location and radius for finding relevant sellers or service providers. The Google Maps API was also instrumental in enabling the visualization of these locations on a map within the application. Additionally
- I integrated the Google Translate API to translate user requests and seller descriptions into English
- ensuring that the application could process and match requests accurately regardless of the original language used. This combination of APIs enabled a seamless
- multilingual
- and location-based service experience for users.
Team
By
Daniel Getahun (BlackDiamond)
From
Ethiopia