Vishvas Foundation Voice Platform

A web platform for spiritual Ashram speech and voice-related tasks. Built with Next.js (App Router) for the designated UI and Python (FastAPI) for advanced audio machine learning tasks.

Features

1. Transcription: Batch transcription of audio archives into text, specialized for Hindi spiritual discourse.
2. Voice Cloning: Synthesis interface (placeholder UI) for various vocal avatars.
3. Audio Quality Improvement: Enhancement and normalization of older recordings using neural noise reduction.
4. Content Indexing: Advanced semantic search (coming soon).

Tech Stack

• Frontend: Next.js, React, Vanilla CSS modules, Custom Design System ("Sacred Luxury")
• Backend (API): Python, FastAPI
• Audio ML: Librosa, Noisereduce, Pydub, Sarvam AI APIs

Setup & Local Development

Prerequisites

• Node.js (v18+)
• Python 3.10+
• SARVAM_API_KEY for transcription features

Running the App

1. Start the Backend

   cd backend
   python3 -m venv venv
   source venv/bin/activate
   pip install -r requirements.txt
   cp .env.example .env # Add your Sarvam API Key
   uvicorn main:app --port 8000

2. Start the Frontend

   npm install
   npm run dev

   The app will be available at http://localhost:3000

Design System

The application employs a custom "Sacred Luxury" aesthetic:

• Typography: Cormorant (Display) & DM Sans (Body)
• Palette: Ivory base, deep charcoal text, warm gold accents.
• Interactions: Subtle, organic fade-ins, breathing hover states, and grain textures that evoke physical archives.