Metra | AI powered live Voice+Video Interviewer

This project is a web-based AI-powered interviewer that conducts screening interviews with candidates. It uses Google's Gemini 2.0 flash live api to ask relevant questions based on a job description and a candidate's resume.

Features

• Dynamic Job Descriptions: Edit the job description to match the role you're hiring for.
• Resume Analysis: Upload a candidate's resume in PDF format. The AI will parse the resume and use it to ask personalized questions.
• AI-Powered Interviews: The AI interviewer, "Mentra," asks questions based on the job description and the provided resume.
• Real-time Transcription: View a real-time transcription of the conversation.
• Tool Use: The AI can use tools, such as ending the call or recording an evaluation.
• Evaluation: The AI provides a detailed evaluation of the candidate's performance at the end of the interview.

Getting Started

Installation

1. Clone the repository:

   git clone https://github.com/your-username/interview-agent.git

2. Navigate to the project directory:

   cd interview-agent

3. Install the dependencies:

   npm install

Running the Application

First add a .env with REACT_APP_GEMINI_API_KEY filled with your gemini api key.

Note: This method of storing the api key on the client side is not secure but since this is a demo, I am not implmenting a separate backend server for this.

To start the application, run the following command:

npm start

This will start the development server and open the application in your default browser at http://localhost:3000.

Project Structure

The project is structured as follows:

• public/: Contains the public assets of the application.
• src/: Contains the source code of the application.
  • components/: Contains the React components.
  • contexts/: Contains the React contexts.
  • hooks/: Contains the custom React hooks.
  • lib/: Contains the libraries and utility functions.
  • assets/: Contains the static assets, such as images and fonts.
• README.md: This file.

How to Use

1. Welcome Screen:

   • Edit the job description in the text area.
   • Upload a candidate's resume in PDF format.
   • Click "Go to Interview Screen" to proceed.

2. Interview Screen:

   • Click the "Start" button to begin the interview.
   • Use your microphone to answer the AI's questions or optionally also use the camera

3. Final Evaluation:

• The AI will end the call and provide an evaluation when it has enough information. You can also say "end call" to stop the interview yourself.

Contributing

Contributions are welcome! Please see the CONTRIBUTING.md file for more information.

License

This project is licensed under the Apache License 2.0. See the LICENSE file for more information.