EchoScribe

AI-powered meeting transcription and summarization.
Upload your audio or video recordings, and get clean transcripts, actionable bullet points, and meeting highlights instantly.
Built with Next.js, Tailwind CSS, and Hugging Face models.

---

🧠 Description

EchoScribe is a minimal, modern web app that converts your meeting recordings into readable transcripts and concise summaries.
It supports audio and video uploads, automatically extracts audio, and leverages AI models for fast, accurate transcription and summarization.

Whether you want to save time after meetings or quickly extract action items, EchoScribe makes it effortless.

---

✨ Features

• Upload audio (MP3, WAV) or video (MP4, MOV) recordings
• Automatic audio extraction from video files using FFmpeg
• AI-powered transcription using Hugging Face Whisper
• AI-powered summarization into clear bullet points
• Real-time upload progress with toast notifications
• Clean, responsive frontend built with Next.js + Tailwind CSS
• Minimalistic dashboard to preview transcripts and summaries

---

⚙️ Workflow

1. Upload file

   • Users upload audio or video files via the frontend.

2. Audio extraction (if video)

   • If the file is a video, FFmpeg converts it into audio (MP3/WAV).

3. Transcription

   • The backend sends the audio file to a Hugging Face Whisper model, which returns a text transcript.

4. Summarization

   • The transcript is processed by a Hugging Face summarization model to generate bullet points and highlights.

5. Display results

   • The frontend displays both the transcript and the summary in a clean, readable format.

---

🧱 Tech Stack

Layer	Technology
Frontend	Next.js, React, Tailwind CSS
Backend	Node.js, Express
AI Models	Hugging Face Inference API (Whisper + BART)
Video/Audio Processing	FFmpeg

---

📁 Project Structure

echoscribe/
│
├── backend/        # Express.js API (transcription + summarization)
│   ├── server.js
│   ├── package.json
│   └── .env.local
│
├── frontend/       # Next.js client app
│   ├── app/
│   ├── components/
│   ├── data/
│   ├── package.json
│
└── README.md

Backend URL https://echoscribe-mqa5.onrender.com Frontend URL https://echo-scribe-delta.vercel.app/