OpenBook

OpenBook is an interactive AI-powered research and study studio (inspired by NotebookLM) that turns your documents, PDFs, and web links into dynamic knowledge artifacts and interactive multi-speaker debate podcasts.

Unlike static audio overviews, OpenBook introduces a Live "Interrupt & Ask" capability: listeners can pause the podcast mid-episode to ask the AI co-hosts questions, receive an in-character audio response grounded in their source notes, and seamlessly resume the show.

---

📺 Watch the Live Demo

Demo Video: https://x.com/swamiabhishek45/status/2090689196923212107?s=20

---

Key Features

• Multi-Host AI Debate Podcasts: Transforms source materials into dynamic, intellectual audio debate shows featuring two AI hosts (Alex, the analytical host, and Jordan, the challenger) powered by ElevenLabs voice synthesis.
• Live "Interrupt & Ask": Pause the audio show at any timestamp to ask specific questions. The hosts generate an immediate conversational response grounded in your workspace documents before continuing the episode.
• Automated Learning Artifacts:
  • Structured Executive Reports & Summaries
  • Interactive Study Quizzes with explanations
  • Revision Flashcards
  • Visual Mind Maps
  • Key Takeaways
• Multi-Format Source Ingestion:
  • PDF document uploads (parsed, chunked, and stored in Cloudinary)
  • Web links and articles scraped via Firecrawl
  • YouTube video transcripts
  • Raw text notes and research snippets
• Grounded Semantic RAG: Vector embeddings and semantic search powered by Pinecone to ensure citations and factual accuracy without hallucinations.
• Event-Driven Background Processing: Asynchronous artifact generation pipeline managed by Inngest.
• Pro Tier & Subscriptions: Integrated Razorpay subscription workflow with usage limits, plan upgrades, and gated premium features.

---

Tech Stack

Frontend (/client)

• Framework: Next.js (App Router), React, TypeScript
• Styling: Tailwind CSS, Radix UI, Lucide Icons
• State & Data Fetching: TanStack React Query
• Audio Playback: Custom HTML5 Audio Player with dynamic seeking and real-time waveform visualizers

Backend (/server)

• Runtime: Node.js, Express, TypeScript
• Database & ORM: PostgreSQL with Prisma ORM
• Authentication: Better-Auth (with Google OAuth & Session Management)
• Vector Database: Pinecone
• Asynchronous Jobs: Inngest
• File & Media Storage: Cloudinary (PDFs & Podcast Audio)
• Payment Gateway: Razorpay

AI & Speech Services

• Voice Synthesis: ElevenLabs REST API
• LLM Reasoning & Extraction: Google Gemini & OpenAI GPT-4o / GPT-4o-mini
• Web Scraping: Firecrawl & Tavily Search

---

Architecture Overview

┌─────────────────────────────────────────────────────────────┐
│                      Next.js Client                         │
│   (Studio Panel, Podcast Viewer, Chat, Workspace Layout)    │
└──────────────┬──────────────────────────────▲───────────────┘
               │                              │
          REST / SSE                     HTTP Audio
               │                              │
┌──────────────▼──────────────────────────────┴───────────────┐
│                    Express Backend Server                   │
│   (Auth, Workspace Routes, Artifact API, Audio Streaming)   │
└──────┬──────────────┬──────────────┬──────────────┬─────────┘
       │              │              │              │
┌──────▼──────┐┌──────▼──────┐┌──────▼──────┐┌──────▼───────┐
│  PostgreSQL ││   Pinecone  ││  Cloudinary ││  ElevenLabs  │
│  (Prisma)   ││  Vector DB  ││ Media Store ││     TTS      │
└─────────────┘└─────────────┘└─────────────┘└──────────────┘
                      ▲
                      │ Triggers Background Jobs
               ┌──────┴──────┐
               │   Inngest   │
               │   Worker    │
               └─────────────┘

---

Getting Started

Prerequisites

• Node.js: v18.18+ or v20+
• PostgreSQL: Local instance or hosted database (e.g. Neon, Supabase, Railway)
• API Keys: OpenAI, Google Gemini, ElevenLabs, Cloudinary, Pinecone, and Firecrawl

---

1. Clone the Repository

git clone https://github.com/your-username/openbook.git
cd openbook

---

2. Backend Setup (/server)

1. Navigate to the server directory:

   cd server

2. Install dependencies:

   npm install

3. Create your environment configuration:

   cp .env.example .env

4. Configure your .env file with your credentials (see Environment Variables).

5. Run database migrations and generate the Prisma client:

   npx prisma db push
   npx prisma generate

6. Start the backend development server and Inngest dev server:

   # Terminal 1: Express Server
   npm run dev
   
   # Terminal 2: Inngest Dev Server
   npm run inngest

---

3. Frontend Setup (/client)

1. Navigate to the client directory:

   cd ../client

2. Install dependencies:

   npm install

3. Create your environment configuration:

   cp .env.example .env.local

4. Start the frontend development server:

   npm run dev

5. Open your browser and navigate to:

   http://localhost:3000

---

Environment Variables

Server (server/.env)

Variable	Description
PORT	Backend server port (default: 8081)
CLIENT_URL	Frontend origin URL (e.g. http://localhost:3000)
DATABASE_URL	PostgreSQL connection string
BETTER_AUTH_SECRET	32+ character random secret for session encryption
BETTER_AUTH_URL	Base server auth URL (http://localhost:8081)
OPENAI_API_KEY	OpenAI API key for text/chat generation
GEMINI_API_KEY	Google Gemini API key
ELEVENLABS_API_KEY	ElevenLabs API key for multi-speaker TTS
PINECONE_API_KEY	Pinecone API key
PINECONE_INDEX	Pinecone index name
CLOUDINARY_CLOUD_NAME	Cloudinary cloud name
CLOUDINARY_API_KEY	Cloudinary API key
CLOUDINARY_API_SECRET	Cloudinary API secret
CLOUDINARY_UPLOAD_PRESET	Cloudinary unsigned preset name
FIRECRAWL_API_KEY	Firecrawl API key for URL scraping
RAZORPAY_KEY_ID	Razorpay Key ID for payments
RAZORPAY_KEY_SECRET	Razorpay Key Secret

Client (client/.env.local)

Variable	Description
NEXT_PUBLIC_API_URL	Public backend API URL (http://localhost:8081)
NEXT_PUBLIC_BETTER_AUTH_URL	Better Auth server URL (http://localhost:8081)

---

Project Structure

openbook/
├── client/                     # Next.js Frontend
│   ├── app/                    # App Router pages & layouts
│   │   ├── (auth)/             # Login, signup & auth pages
│   │   └── (protected)/        # Protected workspace & dashboard pages
│   ├── components/             # Reusable UI components (Radix / Tailwind)
│   ├── features/
│   │   ├── learn/              # Studio panel, podcast viewer, artifacts
│   │   ├── sources/            # File uploads, link ingestion, source manager
│   │   ├── chat/               # Grounded RAG conversational interface
│   │   └── billing/            # Pricing modal & Razorpay checkout
│   └── lib/                    # API client & helper utilities
│
└── server/                     # Express & Prisma Backend
    ├── prisma/                 # Database schema & migrations
    ├── src/
    │   ├── controllers/        # Express route controllers
    │   ├── routes/             # API routing definitions
    │   ├── services/            # Business logic & AI generation pipelines
    │   ├── inngest/            # Event-driven background workers
    │   └── lib/                # Cloudinary, ElevenLabs, Pinecone & Gemini clients
    └── uploads/                # Local file storage fallback

---

Key API Endpoints

• POST /api/workspaces/:workspaceId/artifacts: Generate a learning artifact (Summary, Quiz, Flashcards, Mindmap, Report, Podcast).
• GET /api/audio/:workspaceId/:artifactId: Direct byte-range audio streaming for generated podcasts.
• POST /api/workspaces/:workspaceId/artifacts/:artifactId/podcast/interrupt: Submit a live listener question to the AI podcast hosts.
• POST /api/workspaces/:workspaceId/sources: Ingest a source document (PDF, YouTube link, Web URL, text).
• POST /api/workspaces/:workspaceId/chat: Stream grounded RAG answers with citations.

---

License

This project is licensed under the MIT License.