Jobwise AI

Inspiration

We were inspired to build Jobwise AI to help people prepare for interviews because we know that it can be stressful preparing for interviews. We wanted to create a tool that was easy to use yet meaningful. Thus, we build Jobwise AI.

What it does

Jobwise AI helps people practice job interviews in a realistic and effective way. It generates interview questions tailored to the user’s target job and lets them answer either by typing or speaking. After the user submits their answer, they are given the chance to get feedback from Openai's ChatGPT.

How we built it

We built Jobwise AI using Streamlit for the frontend to create an interactive and user-friendly interface. We also used embedded JavaScript to handle microphone recording. On the backend, we integrated OpenAI’s GPT to generate interview questions and provide personalized feedback, and used the Deepgram API to transcribe spoken answers into text.

Challenges we ran into

While building Jobwise AI, we ran into a few challenges. Getting real-time audio recording to work smoothly in Streamlit was harder than expected and making sure the Deepgram transcription was accurate and fast took some trial and error.

Accomplishments that we're proud of

We’re really proud of what we built in such a short time. When we started building Jobwise AI, we never imagined we could go so far with it. Now, we feel that we have built a very meaningful project that successfully works.

What we learned

Building Jobwise AI taught us about combining multiple technologies such as API keys, libraries and tools to create a program. We learned how to handle audio recording and transcription and how to design a friendly interface.

What's next for Jobwise AI

If we had more time, we would've added additional features to Jobwise AI. We would've added a rating system, where Openai would rate your interview answer out of 10. One thing that could be improved is the transcript quality. When you upload an audio file, the transcript is not perfect and has errors. If we had more time we would try to improve the transcript quality by finding alternatives to Deepgram.

Built With

• deepgram
• javascript
• openai
• python
• streamlit