This repository contains a chatbot built using the OpenAI API, with integrated audio processing and speech recognition via the Google Cloud Speech API. It allows users to capture audio, transcribe it into text, and then generate automatic responses.
Watch the demonstration of the Chatbot here:
- Records audio from the microphone in real-time and saves it as a .wav file.
- Uses Google Cloud Speech API to transcribe the recorded audio into text.
- Sends the transcription to OpenAI's GPT-3.5 model and returns a response based on the provided content.
- Converts the GPT-generated responses into speech using Google Text-to-Speech (gTTS), allowing users to hear the chatbot’s reply.
- openai
- google-cloud-speech
- OpenAI API Key
- Google Cloud Speech credentials (JSON file)
In this project, AJAX is used to create a seamless real-time chat experience. The chat is updated dynamically as messages are exchanged between the user and the bot, without requiring the page to reload.
Consider adding a database to store user conversations for tracking history, user progress, and analytics. Storing interactions could enable personalized responses and suggestions based on previous chats.
To boost performance, especially when handling multiple requests, you can use Celery for background jobs.
- Handling API calls in the background to improve responsiveness
- Enhancing scalability
- Increasing response speed.
- Combined with AJAX, it offers a smoother user experience by providing real-time updates.
- Clone this repository:
git clone https://github.com/narumikat/chatbot-openai.git- Install the dependencies:
pip install - r requirements.txt- Set up credentials:
- Add your OpenAI API Key in the .env file:
OPENAI_API_KEY=your_openai_api_key- Download the Google Cloud service account JSON file for Speech-to-Text API access, and export its path as an environment variable
GOOGLE_CREDENTIALS_PATH=path/to/your-google-credentials.json