The problem Titofy solves

During this quarantine we all are looking for the slightest of interaction. With very less chances of going out , it is improbable to meet new people.
Moreover, the concept of meeting new people through the various meeting apps for today has become bleak.
As it is said, “first impression is the last impression” , half of the meeting apps tend to be a disaster for people who think they do not look good enough & are not socially confident. Nowadays, the factors on which people get matched are still not viable.

This app proposes to solve all these issues. We plan on matching people on the basis of their music taste by checking their playlists.

It is rightly said –
“ You can tell a lot about a person from their playlist”.
Now with this, you can match up with your future forever-after partners & get to know them better. Consider the matching up scheme as an ice-breaker, as the compatibility of songs for both parties will be high.

This is an app based project that aims to match up people of all ages based on their music taste.
It is open for all to use & anyone can fill in their details along with their Spotify playlist. Based on this & using machine learning techniques we find similar registered users to match up with you.

Users can SWIPE LEFT to dislike a person or SWIPE RIGHT to like a person. People who liked each other can then chat with each other.
The people shown in feed are based on a ML model which is done on basis of their Spotify Playlists.

Challenges we ran into

We faced issues in integrating and installing camera plugin while building our app. It was done.

A major issue we faced was deciding on the ML pipeline. We had to decide on a simple clustering algorithm solely because of time constaints. Otherwise we had planned to use a full reccomendation engine model.

Took a while to extract databases. But was done soon. The code for extraction was also quite complex because of the library so a major part of the inital day went in that.

We faced issues in ML hosting. The main issue was compatibility with the versions of all the libraries when hosting our .py files in backend. The issue was resolved after a couple of hours' of hardwork as we had to figure it out by removing & reinstalling all libraries.

APK building was a major issue because of few reasons like OS compatiblity , low end laptop and other issues.
But we together managed to pull together everything and did it.