Skip to main content

Orchestrate AI Agents Up to 5.3x Faster with SPOQ

Coordinate multiple AI coding agents to build software in parallel, with quality gates that catch errors before they ship. See the benchmark results.

Get StartedRead Paper

AI Agent Orchestration Innovations

Multi-agent AI development creates three hard problems: code conflicts, inconsistent quality, and humans locked out of critical decisions. SPOQ solves all three.

Wave-Based Dispatch
Automatically groups independent tasks into parallel waves from your dependency graph. Tasks within each wave execute simultaneously, cutting total completion time by 1.3–5.3x depending on task graph parallelism.
Dual Validation Gates
Score every plan and every deliverable against 10 structured metrics before it ships. Planning gates (90+ threshold) and code gates (80+ threshold) catch errors early, before they compound.
Human-as-an-Agent (HaaA)
You stay in control. SPOQ treats the human developer as a first-class agent who shapes task decomposition, validates plans, and can be consulted by AI agents during execution.

AI Agent Hierarchy: Worker, Reviewer, Investigator

Three specialized tiers, each optimized for its role so you get the right capability at the right cost.

Agent roles and their characteristics in the SPOQ system
TierModelRoleTradeoff
WorkerClaude OpusExecute complex implementation tasks requiring deep reasoning and code generationHighest capability for difficult tasks
ReviewerClaude SonnetValidate completed work against quality metrics with balanced analysisBalanced cost and depth for thorough review
InvestigatorClaude HaikuRapid codebase exploration and information gathering before task executionFast and economical for lightweight tasks

Design Principles

Five engineering constraints that keep multi-agent orchestration predictable and auditable.

  1. Deterministic Quality

    SPOQ uses structured rubrics with explicit criteria to reduce subjectivity in quality assessment. Metric reliability is an area of ongoing research.

  2. Explicit Dependencies

    All task relationships are declared upfront in structured YAML. The orchestrator builds a directed acyclic graph (DAG) to determine safe parallel execution order.

  3. Atomic Task Boundaries

    Each task represents a single, well-defined unit of work. Clear boundaries enable independent validation and prevent scope creep during execution.

  4. Fail-Fast Validation

    Quality checks run immediately after task completion. Failed validations halt the pipeline early, preventing wasted computation on dependent tasks.

  5. Transparent Provenance

    Every decision, validation score, and output traces back to its source. Complete audit trails enable debugging and continuous methodology improvement.

Frequently Asked Questions

Quick answers about how SPOQ orchestrates AI coding agents.

What is SPOQ?
SPOQ (Specialist Orchestrated Queuing) is a free, open-source methodology that coordinates multiple AI coding agents to build software in parallel. Built-in quality gates score every plan and every deliverable before it ships. See the quickstart guide to get started in 5 minutes.
How does SPOQ differ from CrewAI or LangGraph?
SPOQ focuses on deterministic quality validation (two 10-metric scoring gates) and wave-based parallel dispatch. Instead of free-form agent conversations, SPOQ treats orchestration as a scheduling problem with formal dependency graphs and measurable quality thresholds. See the detailed framework comparison for a side-by-side analysis.
What is the 5.3x speedup?
In experimental validation across three case studies, SPOQ achieved speedups ranging from 1.3x to 5.3x compared to sequential single-agent execution. The variation depends on how parallelizable the task graph is. Highly independent tasks see the largest gains. View full benchmark results for detailed performance data.
What AI models does SPOQ support?
SPOQ uses a three-tier model hierarchy: Opus for task execution (Worker), Sonnet for quality assurance review (Reviewer), and Haiku for build failure triage (Investigator). It is designed around the Claude model family via Claude Code. Learn more about each tier on the methodology page.
Is SPOQ free to use?
Yes. SPOQ is open-source under the MIT license. The methodology, skills, and tooling are freely available. You will need a Claude API subscription to run the AI agents. Check our pricing breakdown for estimated costs.
What is Human-as-an-Agent (HaaA)?
HaaA is SPOQ's collaboration model where the human developer participates as a first-class agent, contributing to task decomposition, validating plans, and being available for consultation during execution rather than being excluded from the process. Read more about HaaA integration in the methodology or check the glossary for key terms.