> ## Documentation Index
> Fetch the complete documentation index at: https://koreai.mintlify.app/llms.txt
> Use this file to discover all available pages before exploring further.
# Agent Platform Overview
Build intelligent, autonomous AI agents that understand context, take actions, and deliver results at enterprise scale.
The Agent Platform is a comprehensive solution for enterprises to develop, deploy, and manage AI-powered business applications. It enables the orchestration of intelligent agents across service, work, and process automation while upholding enterprise-grade security and governance.
## Key Components
The Platform's key components provide comprehensive capabilities that empower organizations to efficiently scale and manage AI deployments across the AI lifecycle.
| Component | Description |
| -------------------------------------------------------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ |
| [Multi-Agent Orchestration](/agent-platform/orchestrate) | Define agent behaviors, skills, and integrations through low-code, declarative programming. Give agents advanced reasoning and planning skills to analyze their environment, act effectively, and operate autonomously. Supports real-time streaming across voice and chat channels. |
| [Tools & Integrations](/agent-platform/tools) | Build reusable tools agents invoke to perform tasks or access external systems. Create tools via visual workflow builders, custom JavaScript or Python, pre-built enterprise connectors, or external API integrations. |
| [Knowledge & RAG](/agent-platform/knowledge) | Enable context-aware responses with hybrid retrieval-augmented generation. Supports agentic retrieval where multiple agents collaborate to find information, and adapts to user preferences, current tasks, and interaction history. |
| [Model Hub](/agent-platform/models) | Deploy and manage AI models including fine-tuned domain-specific models. Use any AI model—OpenAI, Anthropic, Google, or custom—and seamlessly integrate open-source models and third-party providers. |
| [Prompt Studio](/agent-platform/prompts) | Design, test, and optimize prompts collaboratively with real-time co-editing, commenting, and version tracking. Compare prompts across models, use variables, and deploy through an integrated prompt library. |
| [Evaluation Studio](/agent-platform/evaluation) | Measure AI performance with AI-driven automated assessments and human evaluator insights. Use customizable scoring rubrics and real-time performance monitoring to tailor evaluation criteria and improve agent quality. |
| [AI Safety & Guardrails](/agent-platform/guardrails) | Ensure responsible agent operations with PII anonymization, input/output guardrails, advanced content filtering, and granular role-based access controls. Maintain compliance with data protection regulations. |
| [Analytics & Observability](/agent-platform/analytics) | Gain deep insights into agent performance through detailed activity capture, session logs, conversation histories, and usage metrics. Includes end-to-end tracing of agent orchestration and streaming of agent thoughts for transparency. |
| [Interaction Context](/agent-platform/memory) | Manage interaction state throughout its entire lifecycle. Maintain continuity across sessions by tracking previous interactions, user preferences, and relevant data to provide contextually appropriate responses. |
| [Agent Protocol](/agent-platform/apis) | Communicate seamlessly between agents, tools, and external systems using a standardized protocol. Supports multi-modal content, real-time streaming, and flexible API design for integration with any external system. |
| [Marketplace](/agent-platform/agents/agentic-apps/marketplace) | Access a comprehensive library of pre-built agents, tools, and extensions across diverse domains and use cases. Accelerate development with customizable templates, connectors, and sample implementations. |
| [Enterprise CI/CD](/agent-platform/deploy) | Manage granular versioning and multi-stage deployments across separate environments (QA, UAT, Prod). Ensure consistent deployments, thorough testing and validation, and a rigorous software development lifecycle for agent-based solutions. |
| [Collaboration & Audit](/agent-platform/deploy) | Facilitate streamlined team collaboration with customizable roles and permissions. Maintain traceability, accountability, and compliance through comprehensive audit logs that track all Platform activities. |
| [SDKs](/agent-platform/sdk-overview) | Build and integrate Platform capabilities with comprehensive SDKs for custom agents, tools, Search/RAG, and observability traces. Available for developers to efficiently extend and embed Platform functionality. |
## How It Works
```mermaid actions={false} theme={null}
flowchart TD
U[User Request] --> O["Orchestrator
Analyzes intent, selects agents, coordinates"]
O --> A[Agent A Specialist]
O --> B[Agent B Specialist]
O --> C[Agent C Specialist]
A --> T1[Tools]
B --> K[Knowledge]
C --> T2[Tools]
T1 --> R[Unified Response]
K --> R
T2 --> R
```
***