Kickback

KICKBACK INPSO

Friends make life better and we want to make your life better! Connect with people with similar interests as you at your university and make new friends faster.

FUNCTION

Kickback uses AI-enhanced and continuous user data-driven algorithms to create the best subset of people for developing strong connections. Using SwiftUI with Stream as a frontend and an Express server communicating with Cohere, Twilio, APNs, and Supabase.

Here is a simple description of the user flow:

Users go through the onboarding process, where Kickback collects your university/year of study, topics of interest, and social links
Each day, the user kicks back for 5 min/session where the user is paired with other people at their university that share complementary interests/traits from user inputed and system analyzed data
At the end of each session, you get a summary where you choose to send your profile to anyone you’d like to keep in contact with. If you both “swipe right” on each other, you get added to each other’s base and can view each others contact info + continue engaging
If you didn’t vibe with someone, no worries, you don’t have to ever talk to them again… even if they swipe on you
Also, it’s just great fun to KICKBACK and relax for 5 min each day chatting to new and old friends

What we Learned

When we were initially trying to map user interests together, we struggled with calculating the proximity between different nodes. Even though we could have utilized a more specialized network for this task, we opted to use Cohere's API to simplify the process by engineering the perfect prompt to get consistent results.
Also, for many simple projects, SwiftUI is a great way to get it spun up quickly, but since there are a lot of limitations on user flow associated with it, should this application go to scale, we would use UIKit to create a more well-built user interface.

Next Steps

Our next steps involve scaling the app to multiple universities one by one, using automatic content moderation and implementing a report feature to prevent harassment and other inappropriate activities, and expanding our relationship grouping systems.