Braven career connect

A backend system and UI platform to connect students with career goals to professionals in those fields at a mock interview event

Comment

Inspiration

Most of us in the team are graduate students and we totally understand how important it is for students to be prepared for interviews. Getting valuable feedback from industry professionals and identifying areas to improve is incredibly important for students beginning their careers. Braven is an inspiring organization that teaches students how to pursue their career goals. The number of students they teach has been rapidly growing in the past year, and they need to update their processes to handle this expansion. One of the challenges they face is running their mock interview events. These events are an opportunity for students to get valuable time with industry experts and practice their interview skills. Braven needs an efficient way to map students to interviewers with the same interests and skillsets for mock interviews. We resonate with this idea and intend to contribute in the best ways we can.

Current Process

In the way the system works today, the people at Braven must individually collect updated interests and skillsets from each student via email. Then they match each student with two unique interviewers, ensuring that the interviewer is scheduled to have only one student for each time slot. While this process was reasonable when they were managing fewer students and professionals, this is not a scalable solution.

What it does

Our solution attempts to streamline the interviewee and interviewer mapping process in order to make it efficient and scalable. The current approach is manual and tedious in terms of time and effort. We are providing the following features:

  1. A mobile-friendly app for students to register for upcoming events based on career interests.
  2. A friendly user-interface for volunteers to mark their availability for upcoming events based on career interests.
  3. The web interfaces contain the profile info section which will capture the interests of both students and volunteers.
  4. Recommender system which will map the students with volunteers based on career interests.
  5. Event matcher logic for checking if the volunteer is available for the event or not.
  6. New career interests can be accommodated by re-training the recommender model.
  7. Scalable solution hosted on the AWS cloud.
  8. Dashboards showing statistics on number of interviews candidates attended, upcoming event slots.

How we built it

We used the following technologies: For Cloud Computing:

  • AWS EC2

For Web Framework:

  • express.js

For Backend:

  • Node.js - for creating APIs
  • MySql - For database storage
  • Flask - For exposing the recommender system's APIs
  • Python cron script - For interacting with Node.js and Flask APIs and updating event matcher data

For Frontend:

  • iOS - for student app
  • React - for volunteer UI

Machine Learning model:

  • Recommender System

Challenges we ran into

Development was stalled for several hours while the PayPal firewall issue was preventing us from connecting to AWS RDS instance.

Accomplishments that we're proud of

  • Reducing a real-world, multi-level problem to a single system solution that provides a scalable design
  • Building the backend and frontend entirely from scratch in such a short period of time
  • Our student-friendly iOS app and website that offer students a beautiful UI experience. The app and website will provide them the ability to explore their career interests and quickly see relevant information about events they are attending
  • Our recommendation model, which uses advanced Machine Learning to offer recommendations for student-professional matches and related career interest to the students

What we learned

We had a great experience and learned a lot in the process. We figured out how to collaborate as a team in an efficient way to come up with an awesome solution using our diverse technical backgrounds in just few hours. We learned how to build up a complex system quickly by splitting up work and leveraging everyones' skills. We learned how to implement and integrate an end-to-end enterprise application in less than one day, build a cloud system that incorporates many individual components, how machine learning models interact with cloud systems, and decreasing the project scope given such a limited time constraint while maximizing all the time that we did have to work on the project.

What's next for Braven career connect

While we accomplished a lot during this Hackathon, there is more we would like to add in the future. We have some great ideas on Natural Language Processing regarding extraction of career interests from user-profiles, connecting the front end to more information on the backend that will reduce the work of the people at Braven and give a better experience for students and professionals.

  • Event recommendation: Based on students' and volunteers' past history, we can gather data on similar users and recommend events for students and volunteers with similar interests.
  • Based on users' information on career interests in the profile section, we can extract the skillset from both students' and volunteers' profiles. We can apply natural language processing (named entity recognition) and auto-fill their career interests. This would further help in increasing the accuracy of the interviewer to interviewee mapping.
Share this project:

Updates