Repository navigation
5. Team Roster
Helped create several front-end interfaces and features. Created a function to allow the user to open the current song directly in Spotify. Designed several interfaces to the standards of our UI/UX mockup including the pre-made playlist screens and worked on the home screen and import screen with Kyuhyun. Curated the pre-made playlists.
Refactored back-end code by designing and implementing two distinct control flows for handling one song as opposed to an entire playlist. Wrote multiple helper functions in order to reuse certain parts of the code between the two control flows, while still handling the different parsing needed for single songs versus playlists. Worked with Franklin, William, and Raymond to debug code and format the JSON into a manner that could be processed by the K-nearest neighbors algorithm. Also worked with them to actually integrate the KNN algorithm (written by Raymond) into our codebase. Kept documentation up to date.
Helped create a python function to parse a Spotify playlist and return only necessary data such as track name/id, artist name/id, image_url (for the album), preview_url (a link to play a preview of the song), and track features including tempo, danceability, and key. Made a simple recommendation function that uses the playlist data returned from the playlist parsing function to give a recommendation based on the features of the playlist. Helped modify functions that take in single tracks as an input parameter to give recommendations for many songs. Created the python function that calls Spotify's search API and returns a list of results to the user. Kept Trello up to date.
Helped created back-end functions that deal with reading in track/playlist information from Spotify's Web API and passing them into the recommender algorithm and to the front-end. Optimized back-end code so as to decrease response time by decreasing the number of API calls and fixed a front-end problem where the back-end is not responding to the front-end. Rewrote and fixed the code for processing the recommendation queue and making request calls in the front end.
Helped create the front-end implementation for the user interface. Created an activity that allows the user to choose between importing from Spotify and selecting a pre-made playlist based on genre. Created an activity that allows the user to input the spotify url to import their playlist and submit it for the recommendation. Helped Franklin and worked with him on making the get request to the backend and receiving the recommendation.
Helped create the recommendation algorithm in the back-end that compares songs features (BPM/tempo, danceability, key, energy and valence) to the song/playlist that the user inputs initially to find the closest matches to the input song(s) which are then returned to the front end. Helped debug functions in the backend that return meta-data and track lists that the recommendation algorithm uses. Helped update the wiki.
Implemented the front-end musical player interface that smoothly streams a song from a preview_url. Implemented the audio feature setting interface that allows the user to easily view the numerical value of a song's audio feature. Implemented the swipe gestures and corresponding song queue behaviors based on the user's feedback. Implemented the final list view to display all songs liked by the user. Helped with the UI/UX flow design and adjust the front-end to make it more visually close to the design.