SmarketPlace

AI listings, secure payments, real-time chat. Buy and sell nearby, with trust and safety. Your ultimate marketplace!

Comment

Inspiration:

The inspiration for this project came into picture when I wanted to get rid of old stuff in my house, such as my working keyboard, mouse and other tiny things. I can always use apps like Craigslist to list out my items, but it's a tedious and hectic process. Hence, I created SmarketPlace where users can sell tiniest of the tiniest things and make money out of old things at home

What it does:

SmarketPlace empowers users to effortlessly list items for sale, connecting with potential buyers nearby. It harnesses Google Vertex AI for streamlined listings, and creating captions, and sorting items into categories for simple search and easy access. The app also provides secure Square payment processing, and facilitates real-time communication.

How we built it:

The tech stack of SmarketPlace involves an iOS app which is completely made with SwiftUI. For Authentication and Database management, I used Google Firebase Firestore. For backend functions to interact with the Google Vertex API, I used simple Javascript and Firebase Functions running as FaaS. I also used Twilio SendGrid to send out emails such as reciepts, refunds and other information.

I also integrated the product with Square API for payments, which I then save to my Firestore DB for simple and faster processing. I trained a model on Google Vertex, and then used Firebase Function to send image to Google Vertex which then calls the trained model and gets the response based of on the trained model. This puts each of the product into various categories which are generated from Vertex AI. I also used the Vertex Image Captioning to generate captions and descriptions for the image.

Challenges we ran into:

The biggest challenge I hit was when I completely depleted my GCP credits very quickly, since Vertex AI is very expensive to execute and run. Hence, I had to re-create another GCP instance with a different email ID and then re-train my model, which lead to consuming a huge chunk of my time, since I had to re-configure the app, as everything including Firestore, Firebase Functions had to be re-created with the new GCP instance.

Accomplishments that we're proud of:

I'm proud to create a MVP, even if I was not able to integrate the chat feature.

What we learned:

During the development of SmarketPlace, I gained valuable insights into AI integration, payment processing, and creating a secure and user-centric marketplace. I also learned the importance of continuous improvement and community-driven features, and how AI plays a huge role in marketing, and making easy seamless works for applications.

What's next for SmarketPlace:

The journey doesn't stop here. In the future, I aim to add more features, enhancing AI recommendations, and fostering a thriving user community. I'm excited about the future of SmarketPlace and its potential to revolutionize online buying and selling.

Built With

Share this project:

Updates