EN

Platform UI is currently English-only. Other languages here track your preference for future translations.

Go to Home

Quarry

AI Knowledge Infrastructure

Document ingestion, semantic search, and intelligent chat interface powered by vector embeddings.

Get Started

Overview

Quarry is a FastAPI-based platform that enables organizations to build intelligent document repositories with conversational AI capabilities. Upload documents, extract knowledge through semantic chunking, and interact with your data through natural language queries.

Key Features

📄

Document Processing

PDF ingestion and intelligent text chunking

🔍

Vector Search

Semantic document retrieval using sentence transformers

💬

Chat Interface

Natural language queries against your document corpus

👥

User Management

Authentication and authorization system

⚙️

Scalable Architecture

PostgreSQL with pgvector extension for vector storage

Tech Stack

  • Backend FastAPI, Python 3.12
  • Database PostgreSQL with pgvector extension
  • Caching Redis
  • ML/AI Sentence Transformers, PyTorch (CPU)
  • Document Processing PyMuPDF for PDF handling
  • Authentication JWT with passlib bcrypt

Quick Start

# Clone and setup
git clone <repository>
cd quarry
cp .env.example .env

# Run with Docker Compose
docker-compose up -d

# API will be available at http://localhost:8000

API Endpoints

/ - Health check and API info
/auth/* - User authentication
/documents/* - Document management
/upload/* - File upload handling
/chat/* - Conversational interface
/health/* - System health monitoring

Architecture

Quarry uses a microservices-ready architecture with clear separation of concerns:

  • API Layer FastAPI routers for different domains
  • Data Layer SQLAlchemy models with PostgreSQL
  • Processing Document chunking and embedding utilities
  • Caching Redis for session and query optimization

How It Works

1

Upload

Ingest PDF documents into the system

2

Chunk & Embed

Intelligent text splitting and vector generation

3

Index

Store embeddings in pgvector for fast retrieval

4

Query

Ask questions via natural language chat

5

Discover

Get precise answers from your document corpus

Start Building Your Knowledge Base

Deploy Quarry and unlock AI-powered document intelligence for your organization.

Get Started View on GitHub

Reconnecting to the server...

Reconnecting in sec.

Failed to reconnect.
The page will reload automatically.

Session paused by the server.

Failed to resume the session.
Reloading the page...