QUANTAF Documentation

Welcome to the official documentation for QUANTAF, the High-Frequency Assurance Engine for Financial Systems.

🚀 Quick Start

Get up and running in seconds:

# Clone the repository
git clone https://github.com/vinipx/QUANTAF.git
cd QUANTAF

# Build the project
./gradlew build

# Run tests
./gradlew test

# Generate Allure reports
./gradlew allureReport

# Serve documentation locally (http://localhost:3000)
./docs.sh

---

📚 Documentation Guide

Core Concepts

• Overview — Purpose, key features, and quick introduction
• Architecture — 4-layer concentric design with detailed class mappings
• Features & Capabilities — Protocol support, AI, reporting, extensibility

Getting Started

• Tech Stack — Dependencies, versions, and library details
• Configuration Guide — Setup, environment variables, YAML structure
• Usage Examples — TestNG, Cucumber BDD, AI-powered test scenarios

Advanced Topics

• Development Reference — Project structure, extension points, testing guidelines
• Allure Reporting — Report generation, enrichment, failure analysis
• CI/CD & Docker — GitHub Actions workflows, Docker Compose, Testcontainers
• Contributing & License — MIT License, contribution guide, code standards

---

✨ Key Highlights

Protocol Support 🌐

• FIX (4.2, 4.4, 5.0) via QuickFIX/J with WireMock-like stubbing
• Message Queues — ActiveMQ Artemis (Jakarta JMS), IBM MQ skeleton
• REST with OAuth2 token management via RestAssured
• ISO 20022 (SWIFT) — AI-powered XML message generation (pacs.008, camt.053, sese.023)

AI & Intelligent Features 🤖

• NLP-to-FIX translation via FixScenarioAgent
• Smart ISO 20022 stub generation via SmartStub
• LLM-powered scenario generation with template fallback
• Pluggable LLM providers (OpenAI, Ollama)

Testing Modes

• TestNG: Direct Java testing with QuantafBaseTest base class
• Cucumber BDD: Gherkin scenarios with OrderStepDefs step definitions

Enterprise Features

• Rich Allure Reports with FIX message attachments and reconciliation details
• Docker Compose for local ActiveMQ Artemis
• Testcontainers for ephemeral CI infrastructure
• GitHub Actions CI/CD with Gradle caching

---

🏗️ Architecture at a Glance

Layer 4: Test Definition (TestNG, Cucumber BDD, Allure Reporting)
    ↓
Layer 3: AI Cortex (FixScenarioAgent, SmartStub, LlmProvider)
    ↓
Layer 2: Logic Core (MarketMaker, TradeLedger, BusinessCalendar)
    ↓
Layer 1: Protocol Adapters (FIX, MQ, REST)
    ↓
Financial Systems (Trading, Settlement, Reconciliation)

Learn more →

---

🎯 Use Cases

• Trading Platform Testing: Validate order execution, fills, and rejections via FIX stubs
• Cross-Source Reconciliation: Three-way comparison of FIX, MQ, and API trade records
• Settlement Verification: T+N business day calculations with NYSE/LSE/TSE calendars
• Realistic Data Generation: Statistical market data using Normal and Poisson distributions
• AI-Driven Scenario Design: Generate order configurations from natural language intent
• Integration Testing: Test across FIX, MQ, and REST protocols in a single test

---

📋 Technology Stack

Component	Version	Purpose
Java	21 (LTS)	Primary language
Gradle	9.3.1	Build automation (Kotlin DSL)
QuickFIX/J	2.3.1	FIX protocol (4.2, 4.4, 5.0)
RestAssured	5.4.0	REST API testing
LangChain4j	0.35.0	LLM integration (OpenAI, Ollama)
Allure	2.27.0	Test reporting
TestNG	7.10.2	Test execution framework
Cucumber	7.18.0	BDD scenario execution
Testcontainers	1.20.0	Ephemeral CI infrastructure

Full details →

---

🔧 Configuration

QUANTAF uses a centralized quantaf.yml with environment variable interpolation via QuantafConfig:

quantaf:
  environment: local

  fix:
    defaultVersion: FIX44
    sessions:
      trader:
        senderCompId: QUANTAF_CLIENT
        targetCompId: TARGET_PLATFORM
        host: localhost
        port: 9876

  ai:
    provider: ollama
    model: llama3
    fallbackToTemplates: true

  ledger:
    amountPrecision: 8
    defaultTolerance: 0.0001

Configuration guide →

---

💡 What's Included

✅ Multi-Protocol Support — FIX, MQ, REST, ISO 20022 in one framework
✅ WireMock-like FIX Stubbing — FixStubRegistry with predicate matching and sequential responses
✅ Cross-Source Reconciliation — Three-way field-by-field comparison with configurable tolerance
✅ AI-Powered Testing — NLP translation and ISO 20022 generation with template fallback
✅ Statistical Data Generation — Gaussian prices, Poisson volumes, Cholesky-correlated series
✅ Rich Allure Reporting — FIX message attachments, reconciliation detail steps
✅ CI/CD Ready — GitHub Actions workflows for build, test, and docs deployment
✅ Docker Support — Docker Compose for ActiveMQ Artemis local development
✅ Extensible — Add custom MessageBroker, LlmProvider, BusinessCalendar implementations
✅ Well Documented — Comprehensive guides with real code examples

---

📄 License

QUANTAF is distributed under the MIT License. See Contributing & License for details.

---

Ready to get started? Begin with Overview → or Jump to Examples →