Skip to main content

intro

IOA Core — Developer Documentation

Welcome to the IOA (Intelligent Orchestration Architecture) Core developer documentation. IOA Core v2.5.2 is an open-source framework for governed AI orchestration — bringing verifiable policy, evidence, and trust to every workflow.

What is IOA Core?​

IOA Core provides a complete governance and orchestration platform for AI systems, featuring:

  • Immutable Audit Chains: Cryptographically verified audit logging with tamper-detection and hash-chained integrity
  • Memory Fabric: Multi-tier storage with hot (in-memory) and cold (persistent) layers, supporting SQLite, S3, and Local JSONL backends
  • 6 LLM Providers: Unified interface for OpenAI, Anthropic, Gemini, DeepSeek, XAI, and Ollama with automatic fallback
  • System Laws Framework: Seven governing principles for AI orchestration enforced at runtime
  • Vendor-Neutral Quorum: Multi-agent consensus with graceful scaling from 1→N providers
  • Working Examples: Six complete examples covering bootstrap, workflows, roundtable, doctor, providers, and Ollama turbo
  • CLI Interface: Command-line interface for system management and development

Quick Start​

  1. Installation: pip install ioa-core
  2. Bootstrap a project: python examples/00_bootstrap/boot_project.py my-ai-system
  3. Run a workflow: python examples/10_workflows/run_workflow.py
  4. Check system health: python examples/30_doctor/doctor_check.py

Note: Examples run offline by default with mock providers. Set IOA_LIVE=1 and configure API keys for live testing.

For complete tutorials, see our Quick Start Guide.

Key Features​

Governance First​

IOA Core is built with governance as a first-class citizen, ensuring all AI systems maintain ethical standards and regulatory compliance through immutable audit chains and System Laws enforcement.

Vendor-Neutral Quorum​

Introduced in v2.5.0, the vendor-neutral quorum policy enables multi-agent consensus with graceful scaling from 1→N providers, featuring sibling model weighting and auditor fallback selection.

Enterprise Ready​

Designed for enterprise environments with features like:

  • Immutable audit logging with cryptographic verification
  • Multi-tenant support
  • Role-based access control
  • Comprehensive audit trails
  • Scalable architecture
  • 6 LLM provider support with unified interface

Compliance Built-in​

Pre-built compliance modules and frameworks for:

  • GDPR (European data protection)
  • CCPA (California privacy law)
  • HIPAA (Healthcare data protection)
  • EU AI Act (European AI governance)

Installation​

Install IOA Core using pip:

pip install ioa-core

Or install from source:

git clone https://github.com/orchintel/ioa-core.git
cd ioa-core
pip install -e ".[dev]"

Getting Started​

Ready to start building with IOA Core? Check out our Quick Start Guide to get up and running in minutes.

System Laws​

IOA Core is built on a foundation of System Laws that ensure ethical AI development and deployment:

  1. Transparency - All AI decisions must be auditable and explainable
  2. Accountability - Clear responsibility and oversight mechanisms
  3. Ethics - Human values and rights are paramount in AI design
  4. Bias Prevention - Active detection and mitigation of discrimination
  5. Privacy - Data sovereignty and protection of personal information
  6. Sustainability - Environmental consciousness and resource efficiency
  7. Safety - AI systems must not cause harm to humans or the environment

Assurance Score​

Monitor your system's compliance and trustworthiness with our Assurance Score metrics.

Recent Updates​

IOA Core v2.5.2 - Latest features include:

  • Immutable audit chains with cryptographic verification
  • Vendor-neutral quorum policy for multi-agent consensus
  • 6 working examples covering all core features
  • Memory Fabric with multiple backend support
  • Improved CLI interface and developer experience
  • Enhanced documentation and tutorials

For more information, visit our company website or GitHub repository.