Mockingbird is an AI-powered interview preparation platform designed to help developers practice both behavioral and technical interview questions in a realistic, interactive environment. By leveraging Google's Gemini 1.5 Flash for intelligence and ElevenLabs for life-like voice interactions, Mockingbird provides a seamless, conversational experience that mimics a real-world technical interview.

• Dynamic Question Generation: Mockingbird parses job descriptions from URLs or text and generates tailored interview questions using Gemini.
• Realistic Voice Interaction: Uses ElevenLabs TTS (Text-to-Speech) and STT (Speech-to-Text) to conduct the interview verbally.
• Technical Coding Environment: Integrated code editor with Pyodide for client-side execution, allowing users to write and run Python code while explaining their logic.
• Intelligent Scoring & Feedback: Provides a comprehensive breakdown of interview performance, including scores for skill alignment, semantic relevance, structure, and delivery.
• Natural Transitions: An AI "interviewer" provides conversational transitions between questions to keep the session engaging.

• HTML5/CSS3: Minimalist, glassmorphism UI.
• JavaScript (Vanilla): Core logic for audio management, state transitions, and API communication.
• CodeMirror: Professional code editor for the technical portion of the interview.
• Pyodide: Enables running Python code directly in the browser.
• Web Speech API: For client-side interim speech-to-text processing.

• FastAPI (Python): Robust and high-performance API backend.
• Google Gemini 1.5 Flash: Used for question generation, technical evaluation, and scoring.
• ElevenLabs API: Powers the "voice" of the interviewer with high-fidelity speech synthesis.
• Sentence Transformers: Used for semantic analysis of interview answers.

• Python 3.10+
• An ElevenLabs API Key
• An OpenRouter (or Google Gemini) API Key

1. Clone the repository:

   git clone [https://github.com/benjaminburnell/mockingbird.git](https://github.com/benjaminburnell/mockingbird.git)
   cd mockingbird

2. Setup the Backend:

   cd backend
   python -m venv venv
   source venv/bin/activate  # On Windows: venv\Scripts\activate
   pip install -r requirements.txt

3. Configure Environment Variables: Create a .env file in the backend/ directory:

   OPENROUTER_API_KEY=your_key_here
   ELEVEN_API_KEY=your_key_here
   VOICE_ID=your_preferred_voice_id

4. Run the Backend:

   python main.py

5. Open the Frontend: Simply open index.html in your browser.

1. Paste a Job Description: Provide a link or text for the role you're practicing for.
2. Speak with the AI: The AI will ask you behavioral questions. Tap the microphone to answer.
3. Solve Technical Challenges: When prompted, use the coding modal to write and test your Python solution.
4. Review Results: At the end of the session, receive detailed feedback on your answers.

Distributed under the MIT License. See LICENSE.txt for more information.

• Benjamin Burnell
• Daknguyen Nguyen
• Graeme Lamain
• Arshiya Shahbazpourtazehkande

Mockingbird: Mastering the art of the interview, one conversation at a time.