Components Overview
Nodes are the fundamental building blocks similar to LEGO, you can build a customized LLM ochestration flow, a chatbot, an agentic app with all the integrations available in Ardor or build your own custom nodes. They represent various microservices with specific functionalities.
Available Categories
Agents
Advanced AI components that can perform complex tasks, make decisions, and interact with various tools
API Chain
Facilitates interaction with various types of APIs using natural language processing
Cache
Efficient mechanisms for storing and retrieving frequently accessed data and LLM responses
Chains
Fundamental components that combine language models, memory, and tools for complex tasks
Document Loaders
Components for extracting and processing data from various file types and sources
Embeddings
Vector representations of text that capture semantic meaning for various NLP tasks
LLMs
Integration components for various large language models designed for conversational AI
Memory
Components for maintaining context and storing information across conversations
Moderation
Tools for filtering and controlling content to ensure compliance and safety
Multi-Agents
Systems where multiple specialized agents work together under coordinated supervision
Output Parsers
Components that structure and format language model outputs into specific data formats
Prompts
Tools for creating and managing structured prompts for language models
Record Managers
Components for tracking and managing document operations in vector databases
Retrievers
Components designed to fetch relevant information from various sources
Sequential Agents
Components for creating complex, multi-step conversational workflows
Speech to Text
Tools for converting spoken language into written text
Text Splitters
Components for breaking down large text documents into manageable chunks
Tools
Various components that enhance AI workflows with specialized functionalities
Utilities
Versatile components for workflow flexibility and control flow management
Vector Stores
Components for storing and querying vector embeddings for semantic search
Key characteristics of nodes:
- Each node is a microservice with its own environment and scaling controls.
- Nodes are stateless and do not store data internally.
- Nodes can be platform-provided or user-created with custom environments and code.
LangChain
Learn how Ardor integrates with the LangChain framework
LangChain is a framework for developing applications powered by language models. It simplifies the process of creating generative AI application, connecting data sources, vectors, memories with LLMs.
Ardor complements LangChain by offering a visual interface. Here, nodes are organized into distinct sections, making it easier to build workflows.
Currently most of ready to use nodes are from LangChain.