Connect aims to eliminate isolation and loneliness by helping you find new groups of friends. Create and join personalized group chats located near you and build genuine connections with people who share the same interests and passions. We have included a map functionality that allows you to see what different group chats are available near you, for you to join with a click of your finger. Additionally, we implemented a trend functionality that allows individuals to see the top group chats people have been participating in within your community.What makes our platform different is its emphasis on the formation of small, tight-knit groups through the "group chat" discovery and creation interface/platform. Additionally, we give users the power to personalize their social interactions with their community, whether that involves how many people you want in your group, what demographic you search for, specific interests, etc. Our users will come to this app looking for genuine, authentic interactions with the hope of building real relationships and intimate communities. Finally, we believe in inclusivity and providing a safe environment, so we strongly utilize Co:here's NLP models to regulate any toxic, discriminatory, or derogatory language.
We used NextJS with React and Node, on top of a CockroachDB cluster in the backend and supported by API calls to Cohere. Co:here's toxicity classification model plays the crucial role of ensuring hate speech and toxic chats are prevented and regulated. We make a classification API call every time a user sends in a chat message, and only allow it through if it is classified as NOT_TOXIC.
We learned a great deal about building RESTful APIs that follow proper practices and in database schema design, as well as engineering database security. It was our first time as well using Vercel and NextJS and we learned how to navigate the full stack technical process of launching a web project using NextJS. Finally, we learned to iterate and rapidly prototype alongside designs, working in unison as a designer and programmer duo.
Connect wants to explore and further gather a user base from local demographics. We want to begin by targeting UC Berkeley students and local residents. Our primary focus is to be able to iterate through the first stages of our design and expose users to this newly used app. We want to take Connect to the next level and create a social impact for those around us. To do so, we value feedback and trial and error to improve this platform and provide the best social experience we can.