WINterview | Devpost

VR
App
App
App

Inspiration

There are a lot of factors that correspond to success during interviews. A candidate may have all the skills but their nervousness or lack of preparedness can easily lose them the job. Interview preparation is extremely important however it is tough to know exactly how to practice. As students who are looking to secure our next exciting coop a trainer would be an excellent tool. That’s why we set out to build Winterview.

What it does

Winterview is a VR interview training solution where users can practice answering questions for specific interviews. Winterview consists of a mobile app where users can sign in and upload job descriptions which the AI will use to prepare suitable questions. The user will then select a job description to interview for and begin an interview with an AI in virtual reality. Data will be collected on various factors such as eye contact, hand gestures, use of filler words and if key words in the job description were used. Users can then use the app to see scores and analytics on their performance.

How We Built It

The Backend was developed using Python and Flask connected to a cockroach db database. Assembly ai was used for audio analytics and gtts for text to speech. The frontend was developed using React and Next.js Charts for the analytics. The Virtual Reality was developed using Unity VR, along with AdHawk Headsets for eye contact tracking.

Challenges we ran into

We initially had a different idea which involved VR Networking, however we only had one Windows computer powerful enough to run the VR headset. This made us lose a lot of time. Another challenge we ran into was getting the wav files to be transferred between the frontend and backend. We had to write the bytes to a file and then load them instead of directly creating the AudioSource from the byte array.