Introducing Interviewly.ai! As we all know the job market is tough, and it is hard to get interviews, so you gotta make the most out of it. This app allows you to paste in the job description, and just like that, you will have generated interview questions. So turn on your webcam and start practicing! You will have constructive feedback, from the tone to the content, everything will be analyzed.
For the tech stack we used Next.js and used Open AI's Whisper API, along with various libraries for audio and video handling. We scratched our idea initially where we were thinking of using Flask and MongoDB ( hence the git repo name haha, we still have ensured to keep interviewly up to show our iteration) however due to some issues we were unable to. This was a great learning experience and we hope this can help someone in the future!
The next step is to explore options for creating a database to store recordings for the user so they can rewatch them and see their progress over time. A few cool ideas are to add sentiment analysis by noticing timestamps of the audio chunks in order to give feedback on when the user is talking too fast or is pausing too much. We are also looking into optimizing our A.I. model more using the OpenAI fine tuning feature very soon! Overall, stay tuned as Interviewly.ai grows into a staple for nailing tech job interviews!
https://github.com/SaqAsh/interviewly2.ai
https://youtu.be/e1gP493V41k