Quill Read Deeper. Think Further. An intelligent EPUB reading companion that doesn't just hold your books, it understands them.

Report Bug · Request Feature · Download Latest Release

Quill reimagines what a reading app can be. Built for readers who want more than a page-turner, Quill combines a beautifully crafted EPUB reader with Aira, an AI reading companion powered by Gemini that can discuss your book, answer questions about the story, and help you understand the text at a deeper level. With access to thousands of free classics via Project Gutenberg and a RAG pipeline that understands narrative context, Quill is the reading experience books deserve.

• Intelligent Query Routing: Aira automatically classifies your question — whether it needs book-specific passage retrieval or can be answered from general literary knowledge — and responds accordingly
• On-Device RAG Pipeline: A hybrid retrieval system combining BGE semantic embeddings (ONNX Runtime) and FTS4 keyword search, merged via Reciprocal Rank Fusion, surfaces the most relevant passages from your book before answering
• Asymmetric Retrieval: For character and event queries, FTS acts as a boolean filter to narrow the vector search corpus, eliminating name-density bias and surfacing the right scenes even in long novels
• Query Expansion: Each question is rewritten into multiple semantic variations with entity extraction and keyword boosting, dramatically improving retrieval precision
• Spoiler Lock: Aira only discusses content you've already read it won't reveal what happens next
• Conversation History: Full multi-turn conversation with history-aware responses, dead-end filtering, and persistent storage per book
• Quick Actions: One-tap prompts to summarize a chapter, identify characters, explore themes, look up words, explain passages, and more

• Full EPUB Support: Clean, distraction-free reading experience with smooth chapter navigation
• Reading Progress Tracking: Tracks your progress per book, per chapter, with percentage completion
• Customizable Experience: Personalize your reading environment with themes and font options
• Resume Anywhere: Instantly jump back to where you left off across sessions

• Thousands of Free Classics: Browse and download public domain books directly from Project Gutenberg
• Popular Feed: Discover trending titles with paginated browsing and automatic sync
• Smart Search: Find books by title, author, or subject across the entire Gutenberg catalog
• Retry & Resilience: Exponential backoff retry logic ensures downloads succeed even on unstable connections

• Cross-Device Sync: Your library, reading progress, and Aira conversations sync seamlessly via Firebase
• Deep Theming: Multiple color themes with OLED black mode support
• Identity System: Custom display name and avatar for a personalized reading profile
• Statistics & Insights: Track your reading habits, session lengths, and books completed over time

Library	Details	Discover	Aira
Reader	Settings	Search	AI

Quill is built on modern Android development standards with a multi-module architecture, ensuring a codebase that is scalable, testable, and maintainable.

Language:             Kotlin
UI Framework:         Jetpack Compose (Material 3 Expressive)
Architecture:         MVVM + Clean Architecture (Multi-Module)
Dependency Injection: Koin
Local Database:       Room (with FTS4 for full-text search)
AI Companion:         Gemini 2.5 Flash (Aira) via Firebase Cloud Functions
Embeddings:           BAAI/bge-small-en-v1.5 (ONNX Runtime, int8 quantized)
RAG Pipeline:         Hybrid Vector + FTS4 with Asymmetric Retrieval
Networking:           Ktor (with exponential backoff retry)
Book Source:          Project Gutenberg
Backend Services:     Firebase (Auth, Firestore, Cloud Functions)
Concurrency:          Kotlin Coroutines & Flow

• Kotlin — Modern, concise, and safe programming language
• Jetpack Compose — Declarative UI toolkit with Material 3 Expressive components
• MVVM + Clean Architecture — Multi-module separation of concerns for maintainable, scalable code
• Room + FTS4 — Robust local data persistence with full-text search indexing across book chunks
• ONNX Runtime — On-device inference for BGE sentence embeddings without server dependency
• Gemini 2.5 Flash — Powers Aira's conversational intelligence and query routing via Firebase Cloud Functions
• Ktor — Lightweight HTTP client for Gutenberg API with retry resilience
• Firebase — Authentication, Firestore sync, and Cloud Functions backend
• Coroutines & Flow — Fully asynchronous pipeline from retrieval to response

• Android Studio (latest version recommended)
• JDK 17 or higher
• Android SDK API 26+

1. Clone the Repository

git clone https://github.com/MohammadAliUstad/Quill.git
cd Quill

2. Firebase Configuration
   • Create a project in the Firebase Console
   • Download the google-services.json file
   • Place the file in the app/ directory of the project
   • Deploy the Cloud Functions from the /functions directory:

cd functions
npm install
firebase deploy --only functions

3. ONNX Model Setup

   • Place the bge-small-en-v1.5.onnx and tokenizer.json files in the assets/ directory of the AI module

4. Build & Run

   • Open the project in Android Studio
   • Sync Gradle files
   • Select your target device/emulator
   • Click Run ▶️

Contributions are what make the open source community such an amazing place to learn, inspire, and create. Any contributions you make are greatly appreciated.

1. Fork the Project
2. Create your Feature Branch (git checkout -b feature/AmazingFeature)
3. Commit your Changes (git commit -m 'Add some AmazingFeature')
4. Push to the Branch (git push origin feature/AmazingFeature)
5. Open a Pull Request

Distributed under the appropriate license. See LICENSE file for more information.

If you encounter any issues or have suggestions for future updates, please open an issue on GitHub or contact the developer directly.

Developer: Mohammad Ali Ustad

Email: Mohammadaliustad@gmail.com

Company: Yugen Tech