SayLess

Table of Contents

1. Overview
2. Features
3. Technologies Used
4. Installation
   • Prerequisites
   • Setup
5. Images

Overview

SayLess is an AI-powered meeting assistant that joins real-time meetings, speaks naturally in a chosen voice, and actively participates in conversations. It can listen, respond when needed, and automatically generate transcripts, summaries, and notes for every meeting.

Features

• Real-time meeting transcription
• customizable AI voice text-to-speech
• AI summary

Technologies Used

• Frontend: TypeScript, React.js, Tailwind CSS
• Backend: Node.js, Express.js, WebSockets (real-time meeting broadcast)
• Database: MongoDB (for user-generated data)
• External APIs:
  • Recall (join meetings)
  • ElevenLabs (text-to-speech)
  • Deepgram (speech-to-text, transcripts)
  • Google Gemini (generating summaries)
• Other: Cloudflare (for CDN), Clerk (for authentication)

Installation

To run the project locally, follow these steps:

Prerequisites

Before installing, ensure you have:

• Node.js
• npm or yarn
• MongoDB account (for database)
• API keys for:
  • Clerk
  • Deepgram
  • ElevenLabs
  • Google Gemini
  • Recall

Setup

1. Clone the repository:

   • git clone
   • cd into repository

2. Set up MongoDB and .env variables in both frontend and backend

3. Install dependencies and run:

   cd backend
   npm install
   npm run dev

   cd frontend
   npm install
   npm run dev
   

Images

System Architecture Diagram