ai-interviewer

LLM-based Survey System with TTS and STT

This project provides a fully working web-based survey system utilizing multiple Large Language Models (LLM), Text-to-Speech (TTS), and Speech-to-Text (STT) technologies, as specified in the Technical Requirements. It is designed to allow administrators to configure models (both via private APIs and downloadable open-source models) easily using a Streamlit admin interface, while end users can participate in surveys through a friendly user interface.

Features

• LLM (Language Models):
  • OpenAI, Claude, Mistral, Yandex LLM, Gigachat, Llama (via Hugging Face)
• TTS (Text-to-Speech):
  • Edge TTS, Google Cloud TTS, Amazon Polly, ElevenLabs, API integrations
• STT (Speech-to-Text):
  • Whisper, Vosk, Google Cloud Speech-to-Text, Amazon Transcribe, API integrations
• SQL Database (SQLite by default)
• Streamlit Interfaces:
  • pages/admin_interface.py — for administration
  • pages/user_interface.py — for end users
• Flexible question generation and editing
• LLM-based scoring and feedback
• Voice and text answers
• Statistics and comparison of LLM vs human scores
• Interactive charts (Plotly)
• Parallel user sessions supported

Requirements

• Python 3.8+
• All dependencies are listed in requirements.txt
• Some models require API keys

Installation

1. Clone the repository (or generate from the creation script).
2. Navigate to the project root directory.
3. Install the required packages:

   pip install -r requirements.txt
   

4. (Optional) Install the package locally:

   pip install .
   

5. (Optional) You may need to download large open-source models (e.g., Llama, Vosk) before first run.

Usage

Admin Interface

streamlit run app.py

The main page will open; select "Admin Interface".

• Admin login is required (ADMIN_PASSWORD environment variable, default is admin).

User Interface

On the main page, select "User Interface".

• The user enters their first and last name, selects a test, and takes the survey.

Project Structure

• src/
  • llm/ — LLM integrations (OpenAI, Claude, Mistral, Yandex, Llama, etc.)
  • tts/ — TTS integrations (Edge, Google, Amazon, ElevenLabs, etc.)
  • stt/ — STT integrations (Whisper, Vosk, Google, Amazon, etc.)
  • utils/ — Utilities (question manager, statistics, audio, etc.)
  • database/ — Database models and handler
  • main.py — Library entry point
• pages/
  • admin_interface.py — Streamlit admin interface
  • user_interface.py — Streamlit user interface
• tests/ — Unit tests
• requirements.txt — Python dependencies
• setup.py — Installation script
• README.md — Project documentation

Contributing

Pull requests are welcome. For major changes, please open an issue first to discuss what you would like to change.

License

MIT