This is a voice-to-speech prototype using [Microsoft's speech-to-text service][https://azure.microsoft.com/en-gb/services/cognitive-services/speech-to-text/#overview]
Please note that this code might not be production-ready. It doesn't include any tests and while the code works, it might have some duplication and the UI is optimized for a phone whose size is 411px by 731px
- Nextjs for the UI layer and API routes
- Tailwindcss for styling
- Azure Speech Congnitive Service for the speech-to-text functionality
- Faker, for simulating and generating some dummy data
- View a list of conversations.
- Search conversations by
sender_name - Start new conversation from "contacts" and search contacts
- View chat with a specific contact and voice-text-functionality
- Simulates a response from your contact
Start by cloning the repo by running git clone https://github.com/joeynimu/talkbox.git. After cloning ensure you have checkout to the specific repo's directory.
You'll need to setup/have an Azure congnitive service. You can setup a free-tier one to begin with. Please see the get started guide here.
Once you have that setup, please copy your SPEECH_KEY and SPEECH_REGION keys from the Azure dashboard. Create a .env.local file and save the values on there. See example below
SPEECH_KEY=SPEECH_KEY_GOES_HERE
SPEECH_REGION=SPEECH_REGION_GOES_HEREOnce you have the variables set, install the dependencies by running yarn install. Then start the dev server by running yarn dev. You can now browse/intereact with the app at http://localhost:3000