Know Everything Ahead of Your Enrollment. Try out product!!!
As course registration is approaching, NYU students across all campuses are struggling to find what courses to take. Albert course search is slow and the students have to either find information about the professor on other website or ask their friends. Another issue is that all the mid-term review and final-review students submit are viewed by professors but not disclosed to other students. As a result, students cannot get fair information about a professor when registering.
What it does
- We are a course/instructor search platform for NYU students to find out the reputation of professors.
- Better UI for seamless course search experience
- Allow students to rate and comment their professors anonymously
- We believe with more student review about professors, NYU students will benefit by enrolling in a class that actually suits them.
How we built it
- We utilized the NYU API (specifically Course Catalog API, Class Roster API, and Faculty API) to match professor names and NetID with classes they are teaching.
- We used Python Flask and bmob.cn NoSQL for back-end development.
- Deployed with Docker and Kubernetes engine for better control of server usage.
- We bought NYU.wiki domain and our service is already deployed on Google App Engine and served by Global Server Load Balancing engine.
Challenges we ran into
- Since we cannot directly use the NYU API to select professors by course filter, we built up a python backend to do aggregation search. Students can either search professors by courses or search professors by their name. Also, we use python backend to gather related professors or courses for a better reference of course selection.
- Mobile devices resolution fragmentation is a big problem for mobile web app designing. In order for better user experience, we use Semantic UI and flex layout to ensure the suitable display for every device. We chose Semantic UI from a number of VUE libraries such as Materialize and iView because of its rich components and elegant aesthetics.
Accomplishments that we're proud of
- Our query speed is much faster than NYU Albert, which is the official course search engine. We implement NYU API to give students a better experience of understanding of professors, course selection, and career decision.
- Both students and professors can use this platform to evaluate both courses and the teaching style of faculties. Although the rating is anonymous, the comment is not, so that it can be ensured that the comments on the Swift Planner are responsible for public use.
What we learned
- Use NoSQL to store simple nonrelational data
- Use mature libraries such as iView, Materialize and Semantic UI to develop frontend systemically.
- We use serverless backend service provided by Bmob (which is the same as
stdlib.com). So that we can just write python code on
Cloud Functionsrather than deploy it on a server. It is more stable so that we do not need to concern about the maintenance and stability of the server. We can leave more time on API / User Interface designing.
What's next for Swift Planner
- We plan to add more course information such as class time and location onto the search.
- Then we can integrate a new course planner feature where students can easily just click a course and it will be mapped on a weekly calendar. Also, we need to add user system to prevent abusive use of rating and comment system.
- Use natural language processing to filter out offensive slandering comments