Skip to content

Speech to text (& translations) using AWS Transcribe & Translate

Notifications You must be signed in to change notification settings

dyte-io/aws-transcribe

Repository files navigation

Dyte's AWS Transcribe Package

Contents

  1. Introduction
  2. About This Repository
  3. Setting Up the Project
  4. Usage
  5. Contributing
  6. License

Introduction

Dyte is a technology company specializing in real-time communication solutions. Our products are designed to enable seamless and efficient video interactions in various applications, ranging from virtual meetings to interactive live streaming. The offerings are known for their robust performance, scalability, and ease of integration.

About This Repository

This package provides plug and play transcriptions & translations.

Prerequisites

This package uses AWS transcribe & AWS translate, which are paid services, provided by amazon.

Therefore, You must have an iam account in your main AWS (Amazon Web Services) account that can deploy the services.

Once done, download the key containing your accessKeyId, secretAccessKey from the iam account for local testing. For production, if deployed using serverless, you may not need any keys.

To keep things simple, in this repo, you will be asked to put keys in .env file. Please alter the code as per the security standards of your application.

Note: Without having accessKeyId & secretAccessKey, this repo may not work, locally.

How to use

  1. Go to server folder
cd server
  1. Replicate .env.example as .env
cp .env.example .env

Open the .env in your choice of Text File Editor and Edit it and Save it.

  1. Install packages
npm install
  1. Run the server
npm run dev

If sucessful, you would see the confirmation in Terminal that it is running on localhost:3001 or the PORT specified in the .env file.

  1. In a new terminal, go to the client folder from root of this repository
cd client
  1. Install packages
npm install
  1. Replicate .env.example as .env
cp .env.example .env

Modify port, if needed.

  1. Run the client
npm run dev
  1. Now go to browser and open localhost:3000 (Change this, if you have modified port in .env of client)

You would see the Dyte meeting loading on this page.

Turn the Mic on and Start Speaking in english (default language) and you should ideally start seeing transcriptions right away.

  1. Go through the client/demo/index.ts & server folder and take whatever is needed to integrate it in your product.

Note: Though this package takes the complexity away from you, We recommend that you put your own security practices & robustness around it and not treat these samples as production-grade copy-paste solutions.

About

Speech to text (& translations) using AWS Transcribe & Translate

Resources

Code of conduct

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published