Reference Architecture

A high-level view of how CoreAI services fit together.

Agentic Friendly

CoreAI is composed of several cooperating services. Some are user-facing, some provide shared AI capabilities, and others support model lifecycle, retrieval, and observability.

Architecture View

CoreAI Domains

DomainMain roleExample components
User experienceProvide the main product entry pointCoreAI Web Portal
API and orchestrationExpose stable application services and AI workflowsCoreAI API, Model Installer
Model lifecycleManage tracked and installable modelsMLflow, Model Installer
LLM access and toolsStandardize model access and MCP integrationLiteLLM, ToolHive
Retrieval and documentsParse content and support semantic retrievalDocling, Milvus
AI observabilityTrace and instrument AI workloadsOpenTelemetry, MLflow

Design Intent

CoreAI keeps AI application capabilities grouped in one layer, while still relying on Runtime for networking, security, storage, inference, and operational services. The goal is to give users and developers stable product-facing entry points even when internal components evolve over time.

On this page