AI Advances

Democratizing access to artificial intelligence

Follow publication

Member-only story

Build Your First AI Voice Chatbot from Scratch with OpenAI (Whisper, GPT-4, TTS), Flask & ReactJS

AI Voice Assistants will be huge in 2024.

Kris Ograbek
AI Advances
Published in
8 min readMar 22, 2024

--

I’m jumping on the bandwagon of AI Voice Chatbots.

They will be huge for many reasons:

  • Speaking is easier than typing.
  • Listening is easier than reading.
  • People are moving from typing to speaking.
  • Typing on mobile devices is time-consuming.

For smartphones, voice assistants will be the default choice!

So I decided to build my first full-stack application to be used purely with voice (check out the repo & the demo). To build it, I chose the package of OpenAI models:

  1. Whisper (speech-to-text). Used to transcribe my voice into text.
  2. GPT-4 or GPT-3.5(Large Language Models). Responses to the transcript from Whisper.
  3. TTS (text-to-speech). It synthesizes the GPT-4 response into audio, such as mp3.

And because my app is working now, I’m here to teach you how you can do the same. We’ll use the following technologies:

  1. Flask for the back-end part.
  2. ReactJS for the front-end part.

--

--

Written by Kris Ograbek

Helping to build AI MVPs | I build with AI & teach about AI | AI Consultant & AI Engineer | My FREE AI course => https://krisograbek.gumroad.com/l/aichatbot

Responses (1)

Write a response