Skip to main content

Architecture

Agent Studio is a core component of the SK NPO platform, designed to facilitate the creation, management, and deployment of AI agents and models. It sits between NPO Service (AI Ops/Workspace) and the underlying NPO Core infrastructure, providing tools and environments to build and manage AI-driven business solutions.

1. Intelligent Agent Development & Operations​

Comprehensive tools for building, managing, and operating AI agents with intelligent decision-making capabilities.

Core Components:
  • Agent Studio: Visual environment for agent design and configuration
  • Workflow: Workflow-based agent orchestration and automation
  • Contextual Cooperator: Multi-agent collaboration and interaction framework

2. Data & AI Infrastructure Orchestration (Data & AI Infra Orchestration)​

Unified platform for managing data, AI models, and computational resources:

Access & Deployment Layer:
  • RAGaaS & Ontology: Retrieval-Augmented Generation and semantic knowledge management
  • GPUaaS Platform: GPU resource management and allocation
  • LLMOps: Language model lifecycle management and optimization
  • Std. I/F (MCP): Standard integration interfaces for external systems
AI Agent Hub:
  • Centralized management of Agents, MCP Servers, and Skills
  • Unified governance and versioning
AI Gateway:
  • API management and routing for AI services
  • Request/response transformation and optimization
Resources & Infrastructure:
  • Data (Knowledge): Knowledge base and vector storage
  • GPU/NPU (Computing): GPU and NPU resource pools for inference and training
  • LLM Model (Open): Open-source language model support and management
  • Service (System): System-level services and APIs

3. AI Observability & Monitoring (AI Observability)​

Comprehensive monitoring and observability for AI operations:

  • Real-time agent performance tracking
  • Model inference metrics and logging
  • System health and resource utilization monitoring