Skip to content

Architecture Layer

The Architecture Layer defines how your AI environment is structured and which technology stacks it uses for services, AI models, data handling, and agentic middleware.

🆕 Agent-to-Agent (a2a) Integration

Agenterprise's key differentiator is built-in Agent-to-Agent (a2a) integration for seamless multi-agent communication. The latest stacks include a2a capabilities:

  • AI Layer (ai-layer-pydanticai): Direct agent-to-agent message passing and coordination
  • Service Layer (service-layer-fastapi-base): Service mesh for distributed agents
  • Middleware Layer (agentic-middleware-layer-redistream): High-performance event streaming between agents

Learn more about available tech stacks.

Available Technology Stacks

Agenterprise provides ready-to-use technology stacks that you can reference in your architecture layer. Choose the stacks that best fit your project requirements:

Service Layer Stacks

Service layers provide the REST API and business logic framework:

  • service-layer-fastapi-base - High-performance FastAPI-based microservice framework with async support
  • Reference: aiurn:techlayer:github:www.github.com:agenterprise:service-layer-fastapi-base
  • Best for: High-performance APIs, real-time applications

AI Layer Stacks

AI layers handle Large Language Model (LLM) integration and agent orchestration:

  • ai-layer-pydanticai - PydanticAI integration for structured AI model interactions
  • Reference: aiurn:techlayer:github:www.github.com:agenterprise:ai-layer-pydanticai
  • Best for: Type-safe AI agent implementations, structured outputs

Data Layer Stacks

Data layers manage data validation, serialization, and persistence:

  • data-layer-pydantic - Pydantic-based data validation and modeling
  • Reference: aiurn:techlayer:github:www.github.com:agenterprise:data-layer-pydantic
  • Best for: Strong data validation, schema definition, JSON serialization

Agentic Middleware Stacks

Middleware layers provide inter-agent communication and event streaming:

  • agentic-middleware-layer-redistream - Redis-based streaming middleware for agent communication
  • Reference: aiurn:techlayer:github:www.github.com:agenterprise:agentic-middleware-layer-redistream
  • Best for: Multi-agent systems, real-time event streaming, distributed agents

Stack Combinations

A typical complete stack setup:

architecture{
    envid = "fb98001a0ce94c44ad091de3d2e78164"
    service-techlayer = aiurn:techlayer:github:www.github.com:agenterprise:service-layer-fastapi-base
    ai-techlayer = aiurn:techlayer:github:www.github.com:agenterprise:ai-layer-pydanticai
    data-techlayer = aiurn:techlayer:github:www.github.com:agenterprise:data-layer-pydantic
    agentic-middleware-techlayer = aiurn:techlayer:github:www.github.com:agenterprise:agentic-middleware-layer-redistream
}

Using Local or Custom Stacks

You can also reference stacks stored locally:

service-techlayer = aiurn:techlayer:local:..:templates:service-layer-fastapi-base
ai-techlayer = aiurn:techlayer:local:..:templates:ai-layer-custom

Discovering More Stacks

Find curated lists for Agenterprise layers at: * Agenterprise AI-Layers List * Agenterprise Service-Layers List

Feel free to: * Clone templates for your own purposes * Create custom stacks by modifying existing templates * Get in contact with Agenterprise to add your custom stacks to the community lists

Architecture Configuration

Overview

 architecture{
    envid = "fb98001a0ce94c44ad091de3d2e78164"
    service-techlayer = aiurn:techlayer:github:www.github.com:agenterprise:service-layer-fastapi-base
    ai-techlayer = aiurn:techlayer:github:www.github.com:agenterprise:ai-layer-pydanticai
    data-techlayer = aiurn:techlayer:github:www.github.com:agenterprise:data-layer-pydantic     

}

Specifications

Attribute Assignment Rule Cardinality Examples
envid %UID% a unique id as %UID% representing the project 1..1 envid = "fb98001a0ce94c44ad091de3d2e78164"
service-techlayer aiurn:techlayer:local:%RELATIVE_LOCAL_PATH%
aiurn:techlayer:github:%GITHUB_DOMAIN%:%PROFILE%:%TEMPLATE%
  • set your path %RELATIVE_LOCAL_PATH% in an urn style
  • %GITHUB_DOMAIN%:%PROFILE%:%TEMPLATE% specifies a public template to a github project in urn style.
 1..1 service-techlayer = aiurn:techlayer:local:..:templates:service-layer-fastapi-base
service-techlayer = aiurn:techlayer:github:www.github.com:agenterprise:service-layer-fastapi-base
ai-techlayer aiurn:techlayer:local:%RELATIVE_LOCAL_PATH%
aiurn:techlayer:github:%GITHUB_DOMAIN%:%PROFILE%:%TEMPLATE%
  • set your path %RELATIVE_LOCAL_PATH% in an urn style
  • %GITHUB_DOMAIN%:%PROFILE%:%TEMPLATE% specifies a public template to a github project in urn style.
 1..1 ai-techlayer = aiurn:techlayer:local:..:templates:ai-layer-pydanticai
service-techlayer = aiurn:techlayer:github:www.github.com:agenterprise:ai-layer-pydanticai
data-techlayer aiurn:techlayer:local:%RELATIVE_LOCAL_PATH%
aiurn:techlayer:github:%GITHUB_DOMAIN%:%PROFILE%:%TEMPLATE%
  • set your path %RELATIVE_LOCAL_PATH% in an urn style
  • %GITHUB_DOMAIN%:%PROFILE%:%TEMPLATE% specifies a public template to a github project in urn style.
 1..1 data-techlayer = aiurn:techlayer:local:..:templates:data-layer-pydantic
service-techlayer = aiurn:techlayer:github:www.github.com:agenterprise:data-layer-pydantic