Nottera 📜✨

A tactile, vintage-inspired academic note-taking environment powered by real-time typewriter dictation and Sarvam AI Saaras v3 multilingual transcription.

---

📖 Overview

Nottera brings the physical elegance of archival books, embossed leather folios, and mechanical typewriters to modern web-based academic note-taking. Designed with a distinct textbook aesthetic, it pairs classic editorial typography and parchment textures with an intelligent dual-engine voice transcription pipeline.

---

✨ Key Features

🎙️ Vintage Cassette Recorder & Typewriter Feed

• Instantaneous Real-Time Dictation: Spoken words appear in real time (<50ms latency) on an emerging vintage typewriter sheet with a mechanical cursor (█).
• Sarvam AI Saaras v3 Engine: High-fidelity Indic and Indian English transcription using Sarvam's state-of-the-art saaras:v3 model.
• Dynamic Mechanical Reel Animation: Rotating cassette reels driven by live audio RMS energy and voice activity detection (VAD).
• Direct Typewriter Keyboard Input: Type directly on the emerging parchment page during recording.

📚 Tactile Academic Folios & Note Library

• Archival Folio Binders: Organize research and lecture notes into custom folios (e.g., Cognitive Science, Macroeconomics, Architecture, Discrete Mathematics).
• Rich Document Reader: Clean serif typography (Source Serif 4), margin notes, paper textures, and bookmark tabs.
• Instant Global Search: Filter across notes, folders, tags, and voice transcripts instantly.
• Quick-Note Scratchpad: Capture thoughts on the fly with persistent auto-saving.

⚡ Resilient Fullstack Architecture

• Vercel Serverless Ready: Native /api/transcribe and /api/speech-to-text endpoints configured for seamless deployment on Vercel.
• Atomic Local Persistence: Local file-backed storage with atomic writes to ensure zero data corruption.
• Export & Backup: Export your complete notebook catalog as a structured JSON archive at any time.

---

🛠️ Tech Stack

Layer	Technology
Frontend	React 19, Vite 8, Vanilla CSS (Design Tokens & HSL Palettes)
Typography	Source Serif 4 (Google Fonts), Courier Prime
Speech-to-Text	Web Speech API (Live stream) + Sarvam AI saaras:v3 (High-accuracy REST)
Backend / API	Node.js Serverless Functions (Vercel) & Standalone Express/HTTP API
Audio Processing	Web Audio API (16 kHz mono PCM resampling, RMS analysis, WAV encoding)
Linting & Testing	Oxlint, Node.js Native Test Runner

---

🚀 Getting Started

Prerequisites

• Node.js 20+ installed
• (Optional) Sarvam AI API Key for Indic speech-to-text

Installation

1. Clone the repository:

   git clone https://github.com/r69shabh/nottera.git
   cd nottera

2. Install dependencies:

   npm install

3. Configure environment variables:

   cp .env.example .env.local

   Add your Sarvam API key in .env.local:

   SARVAM_API_KEY=your_sarvam_api_key_here

4. Start development server:

   npm run dev

   Open http://localhost:5173 in your browser.

---

📡 API Reference

Endpoint	Method	Description
/api/transcribe	POST	Transcribe 16kHz WAV audio using Sarvam Saaras v3
/api/bootstrap	GET	Fetch profile, settings, folders, and saved notes
/api/folders	GET	Retrieve notebook folios and categories
/api/notes?q=...	GET	Global search across notes and voice transcripts
/api/quick-notes	POST	Create a new scratchpad note
/api/voice-memos	GET, POST	List or save transcribed voice memos
/api/settings	GET, PATCH	Read or update user preferences
/api/export	GET	Download full JSON data archive

---

🧪 Development Scripts

# Start Vite client + Node API concurrently
npm run dev

# Run client only
npm run dev:client

# Lint codebase with Oxlint
npm run lint

# Run automated backend test suite
npm test

# Build production bundle
npm run build

# Start production server
npm start

---

🌐 Deployment to Vercel

1. Push your repository to GitHub.
2. Import the project into Vercel.
3. Set the SARVAM_API_KEY Environment Variable in your Vercel project settings.
4. Deploy!

---

📄 License

Distributed under the MIT License. See LICENSE for more information.