/behavioral-interview-agent

Primary LanguagePython

Behavioral Interview Agent API

Overview

The Behavioral Interview Agent API is designed to facilitate interactive, AI-driven interviews by processing audio inputs from users, analyzing the content, and generating responsive audio outputs. It integrates Google Cloud's Speech-to-Text and Text-to-Speech services along with OpenAI's GPT-3.5-turbo to provide a seamless conversational experience.

Features

  • Audio Input Processing: Accepts user's audio inputs and converts them into text for further processing.
  • AI-Driven Interaction: Leverages the GPT-3.5-turbo model from OpenAI to understand the context and content of the user's query and generate an appropriate textual response.
  • Audio Output Generation: Converts the AI-generated textual responses back into audio, facilitating a natural dialogue flow.

API Endpoints

The API comprises two primary endpoints:

1. Version Endpoint

  • Endpoint: GET /version
  • Description: Provides the current version of the API.
  • Response: A plain text containing the version number, e.g., 0.0.1.

2. Respond Endpoint

  • Endpoint: POST /respond
  • Description: Processes an uploaded audio file from the user, generating a corresponding audio response.
  • Request: A multipart/form-data request with an audio file.
  • Response: An audio stream in MP3 format containing the AI-generated response.

Getting Started

To utilize this API, follow the steps outlined below:

Prerequisites

  • Python 3.6+
  • FastAPI
  • Uvicorn
  • Google Cloud Speech-to-Text API
  • Google Cloud Text-to-Speech API
  • OpenAI API (GPT-3.5-turbo)

Installation

  1. Clone the API repository to your local environment.
  2. Navigate to the project directory and set up a Python virtual environment: 
    python -m venv venv
source venv/bin/activate # On Windows use `venv\Scripts\activate`
  3. Install the required dependencies:
pip install fastapi uvicorn python-multipart google-cloud-speech google-cloud-texttospeech openai

Configuration

Google Cloud Services: Set up a Google Cloud project, enable Speech-to-Text and Text-to-Speech APIs, and download your service account key. Set the environment variable GOOGLE_APPLICATION_CREDENTIALS to the path of your service account key file. OpenAI: Obtain an API key from OpenAI and replace "your_openai_api_key" in the application code with your actual API key. Running the API

Start the API server using Uvicorn with the following command:

uvicorn main:API --reload

This will host the API locally, making it accessible via http://localhost:8000.