UM Marketplace (UMM)

Intro
Persona
Unique features
Potential impacts
Homepage
Sell Page with Gemini Smart Post (Auto Fill)
Search with Filters
Product Details
Buying Page (Leave your contact ways here!)
Buyer's Profile Page - "My Sells & Orders" (I want the microwave!)
Seller's Profile Page - "My Sells & Orders" (what a popular microwave!)
Confirmation Page (once you accept one, others will get auto-rejected!)
"Popular Products & Sellers"
Update Product Information (like quantity...)
Buyer's Comment Page (once your order is accepted, it will show on your "My Sells & Orders" Page)
Seller's Comment Section under "My Sells & Orders" Page
Rating System (others can rate the comments!)
Weclome Page & Search with Criteria

Inspiration

Our team at UMICH was inspired by the growing trend of sustainability and the graduation season, where many people post their second-hand goods on social media platforms. We noticed the need for a convenient platform where students and community members could buy and sell second-hand goods easily. This led us to brainstorm and develop our Second-Hand Goods Trading Platform with a future vision of a real campus social media with the magic power of Gemini. Check it out at ummapp.net!

What it does

The UM Marketplace (UMM) is a platform that allows users to buy and sell second-hand goods conveniently through AI-powered smart posts and shopping features. Thanks to our partnership with Google Gemini, we've implemented a smart post feature that allows users to automatically generate content from the media files they upload, enhancing their listings with creative descriptions and tags. In addition, we plan to launch a Gemini Assistant that you with enhanced browsing, buying or selling experiences in a reliable environment.

How we built it

We built the UM Marketplace using Django, a Python-based web framework, for the backend. AWS services such as EC2 for hosting and RDS for database management were utilized for scalability and reliability. Our frontend was developed using HTML, CSS, and JavaScript to create an intuitive user interface and used Django framework. Collaborating closely with Google Gemini, we integrated their generative AI technology into our platform to provide users with a creative playground for content generation.

Challenges we ran into

One of the main challenges was integrating various components of our tech stack seamlessly, including the integration of Web and cloud services with AWS, database management with MySQL, Gemini API and etc. Configuring AWS services and setting up the database required careful planning and coordination. Additionally, ensuring data security and privacy throughout the platform was a priority.

Accomplishments that we're proud of

We're proud to have developed a fully functional Second-Hand marketplace platform that meets the needs of our users. Our platform provides a efficient process of selling second-hand items which could lead to more sales, thus promoting sustainability and reducing waste. The integration of Google Gemini's smart post feature has enhanced the creativity for our users. What's more, auto-fill product information from the uploaded image adds a layer of transparency to the transaction, potentially reducing disputes and improving the overall credibility of our platform.

What we learned

Throughout the project, we gained valuable experience in UI/UX design, web development, database management, and teamwork. We learned how to leverage technologies like AWS, Django, and Google Gemini to build scalable and reliable web applications. Additionally, we honed our skills in frontend development, making the website seamless and easy to use.

What's next for UM Marketplace (UMM)

In the future, we aim to expand the UM Marketplace into a full-fledged social media platform for the campus community. This includes allowing users to record and share videos and audio, creating a more interactive and engaging experience. Additionally, we're exploring the possibility of implementing a Gemini Agent to provide a smart shopping experience. that automates selling tasks or ordering tasks for users through its unique function call features, further enhancing the convenience and usability of our platform, for example, 1) comparing and finding similar products from other online sellers, 2) auto image enhancing etc. To further increase the security level of our platform, we plan to implement third-party login methods such as Google and Duo. This allows users to log in securely using their existing accounts, reducing the risk of unauthorized access and enhancing overall user experience. We are working on a private Github repo, but if you would like to join our team and contribute to the open issues, please feel free to reach out!