atulmy/fullstack-event-driven-architecture

Text-to-Speech and Speech-to-Text application built with a full-stack monorepo, featuring an event-driven, highly scalable architecture using Node.js, React, Redis, MongoDB, and Docker.

Full-Stack Event Driven Architecture

Overview

View in Figjam

Setup

• Prerequisites

  • Node (v18.x)
  • MongoDB
  • Redis

• Clone repository git clone git@github.com:atulmy/fullstack-event-driven-architecture.git speechy

• Switch to directory cd speechy

• Setup

  • Install packages npm install

Running

• Packages
  • common npm start common
  • model npm start model
  • ui npm start ui
• Projects
  • api
    • core
      • npm start api.core
      • localhost:3001
    • stt (speech-to-text)
      • npm start api.stt
      • localhost:3002
    • tts (text-to-speech)
      • npm start api.tts
      • localhost:3003
  • app
    • web
      • npm start app.web
      • localhost:4001
    • console
      • npm start app.console
      • localhost:4002
  • site
    • web
      • npm start site.web
      • localhost:5001

Notes

Redis as a message broker is only used for development purpose. It should be replaced with a more feature rich pub/sub like RabbitMQ, Kafka, Google Cloud Pub/Sub, etc. for production.

Deploying

Check .github/workflow for GCP Cloud Run deployment example. Requires GCP_PROJECT and GCP_SA_KEY added to repository secret along with Redis instance configured in GCP.

License

Copyright (c) 2024 Atul Yadav

GitHub · X (Twitter)

The MIT License