Parallels

Inspiration

We were inspired by MongoDB's vector semantic search feature. In today's world, this type of search is becoming essential, so we wanted to explore it and build a project that highlights its capabilities. As college freshmen, many of us know how challenging it can be to find friends with similar interests. Parallels aims to help by introducing you to people around campus and beyond who share your thoughts and ideals. This could even help students find roommates and build stronger communities.

How we built it

We built a React frontend connected to a backend using MongoDB, Express, and Node.js. MongoDB Atlas vector search powers the core matching experience by comparing user interests and bios.

Accomplishments that we're proud of

We are very proud of creating a graph that shows the relationships between the ten people you have the most in common with. We are also proud of developing an idea that we believe has real potential to make friend-finding easier and more meaningful.

What we learned

We learned how to handle merge conflicts, explored a new tech stack, and made new friends along the way. We also gained a deeper appreciation for MongoDB engineers and for how powerful this database can be when applied effectively.

What's next for Parallels

We plan to add geographic syncing using MongoDB's geolocation features to help users find friends nearby. We also want to add in-site chat so people can connect right away. Since the hackathon was short, our goal was to build an MVP that demonstrates the power of semantic search in MongoDB before expanding into more common features like geolocation.

Built With

• express.js
• mongodb
• node.js
• react
• tailwindcss