Soul of San Franc

Inspiration

The city of San Francisco is becoming increasingly expensive and unsafe, which is causing a lot of people to move to other cities. This is having a rather bleak cascading effect on those who decide to stay back, since they now feel isolated from their friends. From our interviews, we found out that many friend groups have split up because of moving away. To break this isolation bubble, we created a product that is Linkedin, Bumble and Calendly for friend groups!

What it does

To avoid user getting bored of typing his/her/their information repeatedly, our webapp scrapes their existing social media data and find their common interests. Uses vector search to find people and friend groups similar to them in terms of interests and sports, and matches them up. Also makes hang out plans based on availability of the group.

How we built it

The data flow diagram would do a much better job at explaining at https://raw.githubusercontent.com/aryankeluskar/sfhack-backend/master/sfhacks.png

Challenges we ran into

AI and Vector search was difficult, but mentors helped us out.

Accomplishments that we're proud of

Being able to use LLMs so effectively. Also, DBMS by Neurelo is amazing.

What we learned

A lot about LLMs.

What's next for Soul of San Franc

Gamification to encourage further friend group hangouts.

Built With

• fireworks.ai
• google-calendar
• mern
• mongodb
• neurelo
• python
• typescript
• verbwire