đī¸ AI Large Language Models(LLM)
AI LLMs is an enterprise-level large model service gateway that connects to various large model vendors, providing a unified model calling interface that shields API differences between different large model vendors. It handles API key management, load balancing, failover and retry mechanisms, supporting structured output and multi-turn conversations.
đī¸ Vector Database
Vector database is a database for vector storage and retrieval, providing semantic search functionality. It handles vector storage, similarity retrieval, metadata filtering and collection management, supporting various metric algorithms such as cosine similarity, Euclidean distance, etc., providing data support for AI knowledge bases and semantic search.
đī¸ AI Knowledge Base
AI Knowledge Base is an enterprise-level knowledge management system based on RAG (Retrieval-Augmented Generation) technology, supporting intelligent document retrieval and knowledge enhancement functions. It is responsible for document vectorization storage, semantic retrieval, content reranking, and real-time updates, providing precise knowledge retrieval and content management capabilities.
đī¸ AIAgent
AIAgent is the core execution engine for AI applications, implementing reasoning and action loop decision-making based on the ReAct architecture. It is responsible for tool orchestration, dynamic composition and invocation of various business tools and services, maintaining dialogue context, task execution state and data flow state, supporting decomposition and execution of complex business logic, and providing role-based tool access permission management.
đī¸ AI Assistant
AI Assistant is the unified interface for AI application-user interaction, implementing intelligent routing and multi-Agent collaboration based on LangGraph architecture. It is responsible for routing decisions, intelligent dialogue, and workflow control, supporting visual orchestration, complex business logic, and one-click integration capabilities.
đī¸ Agent Events
Agent events are specialized event handling for AIAgent-related operations. They are responsible for monitoring state changes such as Agent creation, execution, and completion, supporting event triggering and processing for key aspects like Agent task startup, progress updates, and result callbacks, providing a unified event handling mechanism for Agent collaboration and task chain execution.
đī¸ Assistant Events
Assistant Events are specialized event handling for AI Assistant interactions. They are responsible for monitoring AI Assistant execution processes, state changes, and user interactions, providing processing capabilities for core interaction events such as user intent recognition, assistant switching, and session management.